SAN FRANCISCO (GenomeWeb) – A number of clinical collaborators testing Karius' cell-free DNA sequencing test for infectious disease reported interim study results and anecdotal evidence of the assay at the Infectious Diseases Society of America's annual IDWeek conference held here last week.
Karius, which spun out from Stanford University in 2014, has been developing technology to isolate cell-free pathogen DNA from blood. It is aiming its technology in particular at finding the cause of sepsis and for diagnosing infection in immunocompromised patients.
The firm runs the laboratory-developed test out of its CLIA-certified and CAP-accredited Redwood City, California-based laboratory, and boasts a fast turnaround time. For samples received before 8:30 am Monday through Saturday, about 80 percent are returned the next day, according to the company. And, as previously reported, the test has a list price of $2,000.
Karius has conducted a study analyzing the clinical diagnostic performance of its test, but has not yet published those results in a peer-reviewed journal. In that study, the firm compared its test with blood culture for 348 patients with suspected sepsis. Overall, the firm found that their test had a sensitivity of 92.9 percent and specificity of 63.0 percent. The company has also not disclosed other performance metrics such as the positive predictive and negative predictive values of its test.
In a presentation at IDWeek, Kathryn Goggin, a clinical investigator at St. Jude Children's Research Hospital in Memphis, Tennessee, reported early results from a study in pediatric oncology patients. Around 18 percent of all such patients develop a bloodstream infection, she said, and infection is particularly dangerous for children with refractory or relapsed disease.
The impetus for evaluating the Karius test was to see whether it would be possible to diagnose infection earlier and even predict whether a patient would develop an infection.
Goggin said that blood samples are routinely analyzed for pediatric oncology patients as part of their clinical care. In some cases this is to determine their complete blood count, and in other cases it is to determine whether feverish patients have an infection.
That made it possible to see whether the Karius test could identify a pathogen from samples that had been collected prior to a patient developing an infection. The researchers compared the Karius test with standard culture-based diagnostics.
In the study, the researchers enrolled patients who had refractory or relapsed ALL or AML. For patients who developed bloodstream infections, standard culture-based diagnostics were carried out but the Karius test was also run on any samples that were available up to seven days before a diagnosis was made up through seven days after the diagnosis. The goal was to to determine whether the Karius test could identify the pathogen before a diagnosis was made by culture.
While the St. Jude team ultimately plans to enroll up to 200 patients in the so-called PREDSEQ study, Goggin presented data from 31 patients. Of those, nine patients developed 11 infections, including two patients who each developed two infections.
The Karius test was able to identify the pathogen that was ultimately diagnosed by culture in nine out of 11 cases. In 10 of the cases, the researchers also had blood samples from the three days prior to culture diagnosis and the Karius test was able to identify the pathogen in eight of those samples.
Goggin discussed a few specific examples. In one, a patient had an Escherichia coli infection and the Karius test was able to identify the pathogen five days earlier than a clinical diagnosis. However, she cited another example of a patient with a polymicrobial infection for whom the Karius test did not identify the causative infections until after a clinical diagnosis was made.
The team also analyzed samples from 16 patients without infections and while the Karius test did identify bacteria and fungi in three of those cases, none of those organisms were associated with bloodstream infections.
Goggin said that in a number of cases Karius identified other microbes aside from the causative pathogen, and said that those findings had "unclear clinical significance." Those cases, she noted, could be polymicrobial infections, contamination, or even a phenomena known as gut bacterial translocation, where indigenous intestinal bacteria invade other tissues and organs and can lead to bloodstream infection. The condition is more common in immunocompromised individuals.
Goggin said that the researchers plan to expand their cohort to between 46 and 100 patients with bloodstream infections and also plan to evaluate the test to quantify the level of pathogen present. If the study ultimately proves successful, Goggin said the goal would be to "hopefully actively monitor high-risk patients."
Other research groups also presented preliminary data from their studies of Karius in immuno-ompromised patients in posters and presentations at IDWeek.
For instance, researchers from the University of Colorado described in a poster presentation how they are analyzing the test for its ability to identify the cause of infection in patients receiving chemotherapy.
The researchers tested Karius in 55 patients, comparing it with blood culture, and the two methods agreed in nine cases. For 31 cases, the Karius test was positive while the blood culture was negative. For 14 cases, both Karius and blood culture were negative, and for one case the blood culture was positive and Karius was negative.
The results were then reviewed by clinical microbiologists to make a final diagnosis and for all 40 positive results from the Karius test they determined that 10 of those calls were possible and the remainder definite or probable. For the 14 cases where both Karius and the blood culture were negative, a clinical diagnosis could not be made in seven.
The researchers also determined that if the Karius results had been obtained in real time, therapy could have been changed for 47 percent of the patients.
A group from Duke University has been evaluating the test for finding the underlying cause of endocarditis, a heart infection. The team analyzed hospitalized patients with endocarditis, many of whom had either a prosthetic valve or an implanted cardiac device. The researchers analyzed 29 patients, 24 of whom had definitive endocarditis. Of those, blood culture was positive for 20 patients. The Karius test identified the causative organism in all 20 of those patients. It also identified pathogens in two out of the four patients who had endocarditis but were culture negative. For four patients who clinicians ultimately determined did not have endocarditis, testing was negative.
It's unclear how Karius will ultimately compare not only with blood culture, but other newer technologies for diagnosing bloodstream infection. While the company has struck up a number of clinical collaborations, studies have thus far been conducted in limited numbers of patients. And, the market for sepsis diagnostics has become increasingly competitive, with a number of companies receiving US Food and Drug Administration clearances for tests last year.
In addition, it will also have to compete with academic laboratories that are developing their own metagenomic sequencing-based tests for infection diagnosis, several of which described such efforts at IDWeek.