A standardized autopsy procurement allows for the comprehensive study of DIPG biology

Abstract

Diffuse intrinsic pontine glioma (DIPG) is one of the least understood and most deadly childhood cancers. Historically, there has been a paucity of DIPG specimens for molecular analysis. However, due to the generous participation of DIPG families in programs for postmortem specimen donation, there has been a recent surge in molecular analysis of newly available tumor specimens. Collaborative efforts to share data and tumor specimens have resulted in rapid discoveries in other pediatric brain tumors, such as medulloblastoma, and therefore have the potential to shed light on the biology of DIPG. Given the generous gift of postmortem tissue donation from DIPG patients, there is a need for standardized postmortem specimen accrual to facilitate rapid and effective multi-institutional molecular studies.We developed and implemented an autopsy protocol for rapid procurement, documenting and storing these specimens. Sixteen autopsies were performed throughout the United States and Canada and processed using a standard protocol and inventory method, including specimen imaging, fixation, snap freezing, orthotopic injection, or preservation. This allowed for comparative clinical and biological studies of rare postmortem DIPG tissue specimens, generation of in vivo and in vitro models of DIPG, and detailed records to facilitate collaborative analysis.

Keywords: autopsy; brainstem glioma; diffuse intrinsic pontine glioma (DIPG); histone 3; orthotopic injection.

Figures

Fig. 1. Postmortem processing of brainstem and cerebellum

Postmortem brainstem and cerebellum were sectioned into axial sections (a) sectioning through the midbrain, pons and medulla (dotted lines) and through the cerebellum. Alternative sections were either fresh frozen (b) or formalin fixed (c). Fixed samples were then cut into sub-blocks, sections were produced and processed for histological staining (d).

Fig. 2. Postmortem processing of whole cortex

Upon removal and processing of the brainstem and cerebellum, whole brain was sectioned into coronal slices (a). Sectioning was performed from frontal (section 1, panel b) through the occipital lobe (section 12, panel b) to generate between 10 to 12 slices and imaged for future reference (b). Alternative sections were either frozen or fixed in formalin.

Fig. 3. Correlation of histological studies with clinical data

Sagittal (a) and axial (b) MR images from a DIPG patient were correlated with autopsied brainstem tissue. The transverse line (a) indicates the plane represented by formalin fixed specimen shown in panel c. Each specimen was clearly labeled (red markings) and cut into sub-blocks. Panel d shows a representative of one sub-block (asterisk in c) that was processed for H&E staining.

Fig. 4. Detailed histological studies of brainstem tissue is enhanced by matched FFPE sections for each frozen specimen

The Cerebral cortex and Brainstem were sectioned and each alternative section was processed as FFPE or fresh frozen. Alternative FFPE specimens were stained for H&E and studied by a pathologist. Histopathology of the frozen section was then judged by the histological readings of the two surrounding FFPE samples. Punch cores were then obtained from the middle frozen section for downstream molecular studies including RNA and DNA analyses.

Fig. 5. Orthotopic injection of DMSO preserved cells results in diffuse pontine tumors

(a) Mice were injected orthotopically in brainstem with cells stored in DMSO. Mice that showed signs of tumor development were sacrificed, brains were fixed and processed for histological studies. H&E staining showed tumor formation in pons (P) as judged by (b) hypercellularity, (c) proliferation marker Ki67, (d) human mitochondrial protein MAB1273 (e) histone 3 K27 trimethylation, and (f) histone 3 K27M mutation.

Fig. 6. Ventricular extended tumor retains characteristics of primary pontine tumor

Histological staining were performed using the primary pontine (a and b) and extended ventricular (c-f) tumor specimens. H&E staining showed highly cellular tumors (WHO IV). Both pontine (b) and extended ventricular (d) tumors were positive for Ki67, (e) while the ventricular retained high expression of histone 3 mutation and (f) reduced histone 3 trimethylation.