When clinicians suspect lung fibrosis and particularly its most devastating form, idiopathic pulmonary fibrosis (IPF), a noninvasive artificial intelligence (AI)-driven digital diagnostic tool may identify subtype classifications facilitating proper treatment at earlier disease stages. On January 16, 2024, the tool, Simultaneously, the American Medical Association adopted relevant CPT [Current Procedural Terminology] billing codes, according to an IMVARIA Inc. press release.
Diagnosis and treatment of the lung inflammation and fibrosis that drive IPF lung function decline are often long delayed, Joshua Reicher, MD, CEO of IMVARIA Inc. and an adjunct clinical professor at Stanford (California) University said in an interview for CHEST Physician.
“There are multiple challenges with this somewhat uncommon condition. Part of the frequent delays in diagnosis is the lack of access to local experts. Another part is vague presenting symptoms like general fatigue, for example, which can have an overlap with a lot of other conditions. The published median average delay in diagnosis after first presenting symptoms is about 2.2 years. But it’s often longer.”
Determining Type of Lung Fibrosis
Conventional diagnosis based on lab tests for inflammatory biomarkers and extensive clinical history is “fairly straightforward,” Dr. Reicher continued, for determining that a patient has some form of lung fibrosis. “The critical element is to find out what type of lung fibrosis and then begin appropriate therapy. The literature lists about 200 different subtypes, but the top 5 make up the majority of cases. The focus with Fibresolve is on improving noninvasive sensitivity, especially for the cases that are less straightforward, but rather indeterminate and therefore particularly challenging,” Dr. Reicher stated.
Will adjunctive diagnostic use of Fibresolve obviate the need for invasive confirmatory tests? Dr. Reicher was cautious. “We like to be thoughtful about our positioning of artificial intelligence and prefer to say that it puts complementary information in the hands of the physician. It’s really up to the clinicians to decide if they have sufficient information to avoid that biopsy.” The uniqueness of Fibresolve, Dr. Reicher pointed out, is that it is widely accessible and does not require hyper-specialized providers. “You can use it at any center that has standard CT scans.”
Reducing Burden on Physicians
An essential feature of Fibresolve use is that its software analysis is conducted centrally. “Part of our goal is to reduce the burden on the clinicians as much as possible, and we try to offload as much of the technical work from them as we can.”
The clinicians send images to IMVARIA Inc. (typically electronically) where they are processed rapidly, and a report is generated with outputs identifying the specific classification, perhaps with one indicating that the findings are suggestive of IPF. Dr. Reicher observed that the Fibresolve’s deep learning algorithm was trained on thousands of cases. “We’re very confident in the results that it puts out,” he said.
“We’re very excited. This is the first FDA-authorized diagnostic tool of any type in lung fibrosis. We really think this supports doctors and patients in areas where there’s a high unmet need,” Dr. Reicher said.
IMVARIA is next developing, in collaboration with the Mayo Clinic, a Fibresolve application for use in lung cancer, he said.