NEW YORK (360Dx) – A malaria test has been developed with an accuracy rate of 97 percent, according to one of its developers, who added that the work to develop it demonstrates that even mature technologies can have new, though not always obvious clinical value.
The work recently nabbed its creators, Oliver Hayden — a professor at the Technical University of Munich, where he holds the Heinz Nixdorf Chair of Biomedical Electronics — and Jan van den Boogaart from Siemens Healthineers, a European Inventor Award.
Neither Hayden nor van den Boogaart have published any data on their work developing the test, but Hayden said that the technology is based on a conventional hematology analyzer, specifically Siemens' Advia 2120 system, a high-throughput system for complete blood count and the differentiation of blood cells.
About seven years ago, he and van den Boogaart began investigating whether such instruments, which Hayden said are routinely used by many clinical labs, could detect malaria in blood samples. Analyzing well-characterized blood samples from malaria patients, as well as samples from healthy patients and those who had infections other than malaria, the two scientists deciphered patterns in what became "the starting point where we could determine that with an algorithm basically combining different information [and] contents of the different blood cells, you can actually get an extremely high sensitivity and specificity for malaria samples," Hayden said.
He and van den Boogaart ran a statistical analysis of the blood parameters of the malaria samples and non-malaria samples, and identified 30 blood values, which were different in malaria patients, compared to healthy ones.
The Advia system has a few "basic physical parameters," including scatter measurements, Hayden explained, and based on different scatter parameters and absorption values, certain pattern information could be gleaned.
"There are certain interdependencies of these variables," and putting these together and performing a statistical method called principal component analysis, "you actually realize there's a certain rule you can derive out of the pattern," Hayden said. He added that most of the information came not from the erythrocytes, as would be expected, but rather from the blood platelets. For example, a shift in the position of the platelets was highly related to the presence of a plasmodium infection.
Based on that finding, he and van den Boogaart developed their algorithm, which can be programmed into a hematology analyzer for detecting malaria.
The algorithm works by taking the physical parameters measured by the hematology analyzer, which are normally used to get a complete blood count. In developing their algorithm, Hayden and van den Boogaart reviewed those parameters and clustered the data, resulting in a malaria infection fingerprint.
Currently, the test remains a research product and the patent and rights to the technology belong to Siemens, Hayden said.
A spokeswoman for Siemens said that commercial availability of the technology is uncertain, depending on results from comprehensive clinical studies, as well as regulatory approvals.
However, according to Hayden, who recently left Siemens to join the Technical University of Munich, an important takeaway from the work developing the algorithm is that even for a mature market and technology, new clinical value can be found.
"A very mature market means usually you have a huge market, costs per test are rather low," Hayden said, adding "in the past there was not a lot going on in," hematology. "Funny enough, all the machines on the market provide such a rich value of information. Only a very small part of this information is actually really used in diagnostics."
But by analyzing the vast amounts of data that come off such analyzers, insights into infectious diseases, such as malaria, and even cancers, can be attained.
"You use existing technology, you use a completely routine test … and you get all that rich information for free," Hayden said.
According to the World Health Organization, there were 214 million cases of malaria in 2015. The same year, there were 430,000 deaths resulting from the disease. Most of the fatalities, 91 percent, occurred in sub-Saharan Africa, but Hayden said the test in its present form would not be appropriate for use in such resource-limited areas because of the infrastructure that would be required for the technology.
Rather, it would be used to test people in Europe, for example, who have visited areas where malaria is endemic, or immigrants from such regions.
"For such conditions, I believe this algorithm would be very interesting because you can screen the patients on the mass level and it fits also very nicely in a central lab environment," Hayden said.