Longitudinal monitoring of diffuse midline glioma using liquid biopsy

Diagnosis of diffuse midline glioma (DMG) portends a grim prognosis, with less than 10% of patients surviving beyond 2 years.¹ Progress in characterizing the genomic and epigenetic landscape of DMG has catalyzed efforts to develop novel therapeutics. Recent work from Cantor et al suggests that blood and cerebrospinal fluid (CSF) liquid biopsy could manifest clinically useful companion diagnostic assays to enable minimally invasive, longitudinal monitoring of responses to emerging DMG therapies.²

Cantor et al demonstrated the feasibility of serial lumbar puncture (LP) and blood collection in a cohort of 17 patients with biopsy-proven H3K27M DMG enrolled in a trial for a novel drug. Cell-free DNA (cfDNA) was isolated from 62 plasma samples and 29 CSF samples, and droplet digital polymerase chain reaction (ddPCR) was performed using primers for H3F3A K27M. H3K27M circulating tumor DNA (ctDNA) was detected in 53/62 plasma samples and H2K27M CSF-tDNA was detected in 28/29 CSF samples, sensitivities consistent with prior reports in DMG.^3–5

A key aspect of this study was the collection of serial biofluid samples—both plasma and CSF. Serial sample collection is critical to enable longitudinal monitoring that can predict therapy response and correlate with disease progression. Cantor et al observed “spikes” in the levels of ctDNA and CSF-tDNA preceding progression in 50% and 45% of cases, respectively. This observation underscores the utility of a quantifiable, molecular biomarker like ctDNA. Standard-of-care radiographic surveillance is unable to offer this level of granularity. Though limited to a handful of cases, the authors also highlight the limitations of radiographic surveillance in DMG. They present cases of pseudo-progression and pseudo-response, both which were correctly adjudicated by ctDNA.

In many ways, DMG represents an ideal use case for liquid biopsy. These are deep-seated, midline lesions for which neurosurgical biopsy carries non-negligible risk. The discovery of recurrent, stereotyped histone mutations^6,7 provides a ready target for off-the-shelf genomic profiling assays. Important work is being performed to standardize DMG ddPCR assays and is requisite if this biomarker is to gain routine clinical use.⁸

There is much neuro-oncology can learn from the evolution of EGFR ctDNA companion assays in lung cancer.⁹ However, the key challenge in designing the rigorous, prospective studies required to determine the utility of ctDNA and CSF-tDNA is the rare nature of DMG. Given the relatively low incidence of DMG and limited overall survival, we advocate that investigators and clinical trial sponsors endeavor to regularly include robust biospecimen banking and strive to collect serial timepoints. This approach may enable the lockstep advancement of novel DMG treatments and the biomarkers to guide the delivery of these therapies.

Acknowledgments

The text is the sole product of the authors and no third party had input or gave support to its writing.

Conflict of interest statement. There are no conflicts relevant to this work to disclose.

References