NEW YORK – Molecular diagnostics company MiraDx has validated its germline microRNA (miRNA)-based Prostox test for predicting late-onset toxicity in response to radiation therapy for prostate cancer.
The Prostox assay, developed by Yale spinoff MiraDx, analyzes a prostate cancer patient's germline DNA for 24 miRNA SNPs (mirSNPs) related to radiation therapy response. MiRNAs are small, noncoding RNA molecules that regulate gene expression. The test is currently available in the US under an early-access program.
Joanne Weidhaas, cofounder of MiraDx, clinical adviser to the company, and professor of radiation oncology at the University of California, Los Angeles, led a study of the assay's diagnostic accuracy, which was recently published in Clinical Cancer Research. As part of that study, she and her colleagues also identified biological pathways involved in different forms of toxicity, opening the door to more personalized radiotherapy and improved patient outcomes.
The study evaluated data from 148 patients enrolled in MIRAGE (NCT 04384770), a Phase III clinical trial comparing magnetic resonance imaging (MRI)-guided to computed tomography (CT)-guided prostate stereotactic body radiotherapy (SBRT).
Specifically, MIRAGE aimed to determine whether aggressive margin reduction with MRI guidance would significantly reduce acute grade 2 or greater genitourinary (GU) toxic effects after prostate SBRT compared with CT guidance.
Weidhaas explained that MRI-guided radiation is the most recent type of such therapy and has reduced short-term toxicities largely by enabling more targeted radiation than had previously been possible by CT guidance.
"But we keep running into radiosensitive people, and we don't know in advance who they are," she said. "If someone is sensitive to radiation [and] is going to form scarring or fibrosis in the prostate, it doesn't matter how carefully you treat only the prostate and not surrounding things, they're still going to have toxicity."
Weidhaas added that to date, no convincing explanation for these different responses has been found within the tumor itself, leading her to speculate that germline genetic variations would better explain these differences.
To evaluate the Prostox signature, the investigators obtained DNA from patients via cheek swab and stratified them into acute, chronic, and late toxicity categories. Acute toxicity was defined as having at least grade 2 toxicity or moderate adverse events within the first three months after radiation treatment. Late toxicity consisted of at least grade 2 toxicity beginning more than three months after treatment, and chronic toxicity meant having toxicity of grade 2 or more that persisted for six months or later posttreatment.
Although the sensitivity of Prostox was lower in this study than in the original study cohort (approximately 58 percent versus 71 percent), it predicted late GU toxicity with an overall area under the curve (AUC), a measure of diagnostic accuracy, of 0.762. A comparison of treatment arms showed that it performed slightly better in the MRI-guided group (AUC 0.83) than in the CT-guided group (AUC 0.71). Additionally, higher Prostox scores correlated with higher degrees of toxicity among both treatment groups.
Importantly, Prostox significantly outperformed clinical variables previously suggested to predict late toxicity. These variables appeared uncorrelated with Prostox values and did not improve its predictive significance, suggesting that Prostox predicts GU toxicity and does not simply reflect preexisting clinical factors that might influence toxicity.
The investigators also identified a germline genetic signature of acute toxicity among the Prostox SNPs, the most important of which appeared to be three SNPs in the MSH2, P2RX7, and TGFB1 genes.
A distinct signature of chronic toxicity emerged, also dominated by a combination of clinical factors and SNPs in the BMP2 and IL1A genes.
"I think this is the first time we've really genetically separated these groups," Weidhaas said.
The researchers also looked for biological pathways that underpin each type of toxicity. Acute toxicity mainly involves pathways related to gene regulation and DNA damage repair such as primary miRNA processing, mitotic DNA damage checkpoint signaling, and regulation of nucleocytoplasmic transport. Chronic toxicity, meanwhile, is enriched for pathways related to apoptosis and cell cycle regulation.
The late toxicity phenotype identified by Prostox consists largely of pathways related to immune cell proliferation and cytokine production. Patients experiencing this type of toxicity typically display significant immune system dysregulation, which may influence the fibrosis, or scarring, often found in their prostate as a result of prolonged inflammation.
Zuzanna Nowicka, a graduate student of radiation oncology at Poland's Medical University of Łódź, who has authored studies of miRNA as cancer biomarkers, praised Weidhaas' study for addressing an unmet need in cancer radiation therapy.
"Genetic predisposition to increased toxicity after radiotherapy is not routinely evaluated in the clinic and may offer important additional information on top of clinical and treatment-related factors," she said in an email, noting that information contained in noncoding sequences such as miRNAs is particularly understudied.
Nowicka called the study's independent validation of previously developed toxicity biomarkers a "definite strength." Prostox's relatively low sensitivity indicated a need for further calibration across different clinical contexts and toxicity scoring systems before the score can be reliably applied in practice.
She also noted that it would be important to understand the mechanisms by which the SNPs included in Prostox translate to a greater risk for late radiation toxicity.
Weidhaas said more studies are in the works. MiraDx is collaborating on the European PACE-A clinical study, for example, which is evaluating radical prostatectomy versus stereotactic radiotherapy for clinically localized prostate cancer, as well as on several US-based cancer trials, which she said would help to broaden the genetic diversity underpinning research into the Prostox biomarker.
In addition, the company is studying toxicity to radiation in other cancer types and to immune therapy, she added.
Scientists have been investigating the use of miRNAs as biomarkers of radiotoxicity for several years, but MiraDx seems to be the first to develop a commercial test.
Prostox is currently available under an early-access program based out of MiraDx's CLIA-certified and CAP-accredited lab in Los Angeles. Weidhaas said the lab is also seeking approval under the New York State Department of Health Clinical Laboratory Evaluation Program and expects Prostox to become available to the wider physician community early this summer.
Other companies seeking to commercialize miRNA-based biomarkers have mainly been using them for screening and early diagnosis. Molecular testing startup MiRoncol, for instance, published a study last year that evaluated a four-miRNA multi-cancer early detection (MCED) assay in 13 tumor types. Singapore-based Mirxes is also developing an miRNA test, for which it recently obtained regulatory approval in Singapore for gastric cancer screening.
A test like Prostox, Weidhaas said, "has been sorely needed in prostate cancer. So we're excited about it."