By Deborah Borfitz
October 13, 2021 | Researchers at Washington University (WashU) are exploring use of cell-free DNA (cfDNA) as a liquid biopsy analyte for solid tumors and, most recently, to track and predict organ damage causes by sepsis. Particularly noteworthy is their liquid biopsy for distinguishing between benign and malignant tumors in a heritable cancer predisposition syndrome, as reported in a groundbreaking study that even National Cancer Institute (NCI) Director Ned Sharpless tweeted about, says WashU radiation oncologist Aadel A. Chaudhuri, M.D., Ph.D.
For that study, which recently published in PLOS Medicine (DOI: 10.1371/journal.pmed.1003734), the team used a blood biopsy to noninvasively diagnose malignant peripheral nerve sheath tumors (MPNST) as distinct from its benign precursor plexiform neurofibroma (PN). By analyzing blood plasma for fragment size and copy number alterations of cfDNA using ultra-low-pass whole genome sequencing, the researchers derived tumor fraction levels correlating with MPNST tumor burden and treatment response.
Their noninvasive classifier was found to differentiate MPNST from PN with 75% sensitivity at the pretreatment phase and 83% with serial analysis, Chaudhuri reports.
An aggressive soft tissue sarcoma, MPNST is the leading cause of mortality for patients with neurofibromatosis type 1 (NF1) cancer predisposition syndrome, says Chaudhuri. Biopsies are particularly dangerous for patients with plexiform nerve tumors because lesions sit on the nerve lining itself, he says, but all the imaging alternatives—including PET, CT, and whole-body MRI—tend to be sensitive but not specific so findings can be quite similar for both cancerous and non-cancerous tumors.
Given the specificity level of the new liquid biopsy test, it could be used to confirm a suspicious imaging finding, Chaudhuri says. It could thereby improve the early cancer detection and monitoring in predisposed individuals.
A cfDNA liquid biopsy test would be particularly valuable for pediatric patients who are hereditarily at risk for cancers, since there isn’t a great way to monitor them currently, says Chaudhuri, who treats patients at Siteman Cancer Center at Barnes-Jewish Hospital and WashU School of Medicine. Physicians dislike doing CT imaging on children due to the risks associated with radiation exposure, he offers as an example, and standard blood tests generally aren’t a good indicator if cancer is present.
The combination of fragmentomics (analysis of the structural properties of cfDNA molecules) with copy number analysis is what enabled Chaudhuri’s liquid biopsy approach to classify NF1 patients as having or not having MPNST. A shorter fragmentation profile correlated with cancer relative to patients with benign neurofibromas and healthy controls.
Urine-Based Methods
Chaudhuri says he began working on liquid biopsies as a chief resident and postdoctoral fellow at Stanford and his interest in the space continued when he came to WashU three years ago. He previously worked with clinical colleagues to publish on the potential of cfDNA tests to detect lung, esophageal, prostate, and colon cancers.
The recent study on nerve sheath tumors published simultaneously, in the same publication, with another on a urine biopsy for detecting minimal residual disease in muscle-invasive bladder cancer (DOI: 10.1371/journal.pmed.1003732). The standard treatment here is radical cystectomy, an extensive surgery that involves removal of the bladder and nearby organs. But many of these surgeries could potentially have been averted altogether based on urine biopsy results showing patients had achieved a complete response to chemotherapy.
By inferring tumor mutational burden from urine samples, the researchers were also able to “precisely identify” the 58% of patients with detected residual disease who might be candidates for adjuvant treatment with immune checkpoint blockade, Chaudhuri says.
To identify residual tumor DNA, researchers used urine Cancer Personalized Profiling by Deep Sequencing (uCAPP-Seq), a next-generation sequencing method developed by Chaudhuri’s mentors at Stanford. CAPP-Seq was first implemented for non-small-cell lung cancer, as initially described in 2014 (DOI: 10.1038/nm.3519) and again, with integrated digital error suppression, in 2016 (DOI: 10.1038/nbt.3520).
Some of the other urine-based approaches currently under development are also “quite promising,” says Chaudhuri. Patients otherwise must undergo an invasive cystoscopy at regular intervals, an uncomfortable and expensive procedure where the urologist inserts a scope into the bladder via the urethra.
In addition to the novel cfDNA method, other urine-based methods being investigated variably look at sediment cells, proteins, and mRNA, as described in a review published by Chaudhuri last year in JCO Precision Oncology (DOI: 10.1200/PO.20.00060). “Ultimately it could be an amalgam of these methods that is most sensitive and specific for detecting residual disease rather than any single method alone,” Chaudhuri says.
Multiomics Approach
With both cfDNA-based liquid biopsies, roughly 20% of patients would potentially be missed and forego treatment, which is “not acceptable from a clinical standpoint,” says Chaudhuri. “It’s hard to make treatment decisions confidently if sensitivity is not greater than 95%,” given that the potential risk of a false-negative result is death from cancer.
If residual disease is missed after chemotherapy and/or radiation treatment and patients skip surgery, they could develop metastatic relapse and potentially die, he says. Moreover, mortality rates for both MPNST and muscle-invasive bladder cancer are unacceptably high, with diagnostic delays worsening the prognosis.
One approach being taken by WashU researchers to improve test sensitivity is to look at cfDNA through multiomics to, for example, detect genomic, epigenetic, and fragment size alterations, Chaudhuri says. In collaboration with the NCI, the liquid biopsy approach for MPNST detection is also going to be tested more rigorously in a multi-institutional cohort. The next big question, he says, is if it can be used to “screen for cancer early” in patients with NF1 cancer predisposition syndrome.
Chaudhuri and his collaborators have filed patents related to cancer biomarkers. Chaudhuri is also co-founder of Droplet Biosciences, an unrelated company in the liquid biopsy space that is focused on identifying residual disease through novel biofluid analyses.
The longer-term plan is to use cfDNA as a biomarker for sepsis, which is the leading cause of hospital death and responsible for about 20% of all deaths worldwide. “Like in cancer, dynamic levels of cell-free DNA features likely correlate with disease status and prognosis,” he says.
The hypothesis is that the research team can develop a liquid biopsy test to predict and monitor organ damage in sepsis patients, which is the primary cause of mortality. “This is especially an issue with the COVID crisis and ICUs filling up,” Chaudhuri notes.
The idea struck him while following the COVID news last year. The National Institutes of Health subsequently funded his roughly $2 million grant proposal for innovative cell-free DNA research in sepsis.
That study is now underway but won’t conclude until 2026—by which time the current pandemic, presumably, will also be over. Still, Chaudhuri says, the principles learned will be important in “paving the way for more precise earlier management of this deadly disease.”