NEW YORK – Monitoring the health of a donated organ, such as a kidney, can be enhanced by looking for cell-free DNA from the donor shed into the recipient's bloodstream. Although results are often reported as a proportion of donor DNA relative to recipient DNA, Natera recently reported that host infection and having organs from more than one donor can change these levels.
The San Carlos, California-based firm is now developing methods to correct for these changes and further refine its Prospera test to monitor for kidney transplant rejection.
Natera collaborated with researchers at the University of California, San Francisco in 2016 to develop its kidney transplant monitoring test, and Medicare contractor Palmetto GBA issued a draft local coverage determination for Prospera in 2019.
Much like the firm's flagship noninvasive prenatal test which looks for the fraction of fetal DNA in a mother's bloodstream, the kidney transplant test uses massively multiplexed PCR and sequencing to interrogate approximately 14,000 SNPs and determine a donor fraction of cfDNA, expressed as a proportion of the genetically distinct recipient DNA.
Phil Gauthier, medical director of the organ health division at Natera and a general nephrologist, said in an interview that there are a number of circumstances and conditions that the firm has recently discovered which can change the donor or host levels, essentially distorting either the numerator or denominator of this fraction.
"We've come to understand the significance of what we are calling 'background' cell-free DNA," Guathier said. "When these tests were originally developed, a [cutoff] percentage was chosen because it was assumed that the variable in the equation would be the donor cfDNA," he said. Essentially, it was presumed that an injured donor kidney would release increasing amounts of cfDNA, and the fraction would increase while the background, or recipient cfDNA, would remain constant.
"It turns out, that's not true," Gauthier said.
Gauthier and his colleagues at Natera authored a series of posters presented at the recent American Society of Nephrology Kidney Week virtual meeting describing cases where background recipient cfDNA varied significantly due to viral infection. Conversely, they also described increased cfDNA due to multiple donors.
The new data is "probably the first public presentation of the fact that background levels can vary so significantly and can have profound implications on how the test performs," Gauthier said.
The level of recipient DNA — essentially the denominator of the transplant monitoring test fraction — can fluctuate. It is especially relevant in this patient population because they are typically given immunosuppressing drugs to help prevent transplant rejection, making them more susceptible to viral infections.
In one case study, the team found that a severe viral infection with COVID-19 resulted in a 57-fold increase in host DNA levels. These extremely elevated levels have the potential to depress the donor fraction reported, and test results could potentially be more accurate if these elevated levels are taken into account.
In the study, a 50-year-old woman developed COVID-19 infection 11 months after a kidney transplant. She developed acute kidney injury and sepsis requiring a type of treatment called renal replacement therapy, which includes interventions such as dialysis. She was also administered the Natera Prospera test five times over a period of about six weeks to monitor for kidney rejection.
According to the ASN poster, the first test was given at day 23 after hospital admission, at which time the patient had 57 times her ultimate stable baseline level of cfDNA attained upon recovery.
In another poster, Natera scientists and collaborators performed testing on three kidney transplant patients after viral infections.
One patient had BK virus infection seven months after transplant, followed at 14 months post-transplant by herpes simplex virus and cytomegalovirus infection of the esophagus. His test revealed .38 percent donor-derived cfDNA — below the 1 percent cutoff for increased risk of active rejection — but the background DNA was found to be 21 times higher than the median level in a separate large study. A biopsy of the donated kidney confirmed acute T cell-mediated rejection.
The second patient had a donor-derived cfDNA fraction of .28 percent, but sevenfold elevated background cfDNA. A biopsy discovered BK virus-associated nephropathy and acute T cell-mediated rejection.
The third patient got Dengue fever a month after her transplant. At six months, her biopsy indicated active antibody-mediated rejection, yet her Prospera test result showed .16 percent donor-derived cfDNA. However, her background cfDNA was 13 times higher than a standard median value.
The authors suggested that serial monitoring of kidney transplant recipients may be important during and after viral infections in order to properly interpret cfDNA measurements.
Gauthier explained that cfDNA is part of the immune response and can be actively secreted by cells, and it is also associated with cellular injury such as that caused by severe systemic infection. In addition to viral infections, other conditions like cancer and obesity could also increase a host's level of cell-free DNA. In addition, Natera recently presented data at The Transplantation Society's annual meeting showing that muscle injury, such as occurs after surgery, can elevate host cfDNA fivefold.
This phenomenon is known to occur in liquid biopsy and non-invasive prenatal testing as well, Gauthier said. For example, one review published this year found an increasing no‐call rate for NIPT tests — in which the test fails to detect a significant fetal fraction relative to the maternal fraction — with increasing maternal obesity.
Overall, the results "have significantly changed our understanding of how this test performs," Gauthier said. In particular, the result has "drawn attention to the fact that patients with various infections can have very elevated backgrounds, which then suppress the donor fraction and can lead to a false-negative test even when rejection is present."
As a result of these background findings, Natera is aiming to improve its Prospera test to integrate background levels into the test results. The firm launched a large multi-center trial called the Prospera Test Enhancement by Detecting Background Cell-Free DNA Levels, or PEDAL, which aims to enroll 500 patients at several transplant centers.
Specifically, the trial aims to improve the sensitivity and specificity of the test by expressing Prospera results "not just as a fraction, but as a fraction that is somehow adjusted for the background level," Gauthier said. The potential adjustments include expressing results as a percentage that is adjusted for background levels, or multiplying the percentage of donor-derived cfDNA by the background to derive the raw amount of donor cfDNA.
Natera is also untangling data from repeat transplantations, in which there can be more than one source of donor DNA in a sample.
It is not commonly known that patients who receive kidney transplants do not typically have their own kidneys removed, Guathier said. Instead, host kidneys are left in situ, and the donor kidney is placed in the lower pelvis, such that a patient typically has three kidneys after the operation.
If a first transplant fails, a second kidney transplant is sometimes warranted, and placed on the other side of the pelvis, Gauthier said.
Approximately 10 percent of all kidney transplants fit this description, and in this scenario a patient winds up with four kidneys, two of which can potentially shed the cfDNA of two different donors.
Natera presented data at ASN from a small study showing that the overall levels of donor-derived cfDNA are slightly higher in patients who are stable after a second transplant, but the result was not statistically significant. Nevertheless, those who experienced rejection of the second kidney still had significantly elevated cfDNA compared to patients who were not experiencing rejection.
"It has always been a point of contention whether a test like this would perform in patients who had more than one source of non-self cfDNA," Gauthier said.
Similar situations can also occur when a patient gets organ donations from multiple donors, such as a kidney from one donor and a liver from another.
A recent estimate placed the market opportunity for Natera's kidney transplant monitoring at $2 billion, with 20,000 new kidney transplants per year and more than 200,000 patients living with a transplant.
The competition in the cfDNA kidney transplant monitoring space includes CareDx's Allosure. For other solid organs, TAI Diagnostics has a heart transplant rejection test called MyTAI Heart, and the firm has said it intends to apply its technology to kidney transplants as well. Transplant Genomics, a firm that was recently acquired by Eurofins, has a test called TruGraf for kidneys, but that test uses blood-based biomarker detection rather than donor-derived cfDNA levels to detect rejection.
Gauthier said that Natera's competitive advantage comes from the high number of SNPs it detects with its tests as well as its extensive experience in non-invasive prenatal testing.
"The way that we fractionate the donor fraction in transplant is exactly the same as we do in prenatal [testing], and since 2009 we've done over 2 million cell-free DNA test, and will probably pass 3 million this year," he said.
The company also is beginning to capitalize on the synergies between oncology and transplant monitoring, Gauthier said, adding, "There is a big interaction between oncology and transplant."
Patients who get tissue transplants are at increased risk for de novo and recurrent cancer because of the immune suppression regimen they are required to undergo. There are also cancers that can lead to the need for transplant.
The firm has formed a committee called NITO, or Natera Innovation in Transplant and Oncology, and will launch three multi-center trials to leverage its oncology test, Signatera, to help transplant patients, and leverage its Prospera transplant test to help patients with cancer.
The three studies are named Colorectal Neoplasm in Candidates Enlisting for Renal Transplantation study (CONCERT), Signatera in Liver Cancer (SIGNAL), and Prospera in Renal Allograft Recipients with Cancer (PARC) study.
"There is no other cell-free DNA company that has products in both transplant and oncology, and a lot of transplant people are very interested in this," Gauthier said.
Gauthier said that any test measuring non-host DNA expressed as a percentage of host DNA might have similar issues. "What's different about our test is that we have figured out a way to measure the background levels on every specimen without changing the workflow," he said.