This article has been updated from a version posted April 6 to clarify the specificity and sensitivity of Veracyte's Percepta Nasal Swab.
NEW YORK – LungLife AI, formerly Cynvenio Biosystems, is recruiting participants for a pivotal clinical study of its LungLB diagnostic assay for evaluating whether indeterminate nodules in the lungs detected on CT scans are cancerous or benign.
This week, the California-based company also reported its financial results for 2022. The firm, which had 14 full-time employees at the end of last year, generated $24,000 in revenues in 2022, from royalty payments from a licensee in China, compared to $195,000 in 2021. Its net loss for the year totaled $7.6 million, or $.30 per share, compared to a net loss of $7.4 million, or $.47 per share, in 2021. The firm ended 2022 with $3.1 million cash and cash equivalents and $4.9 million in short-term deposits.
The LungLB assay, currently offered as a lab-developed test (LDT) out of the company's CLIA-certified lab in Thousand Oaks, California, enriches circulating tumor cells (CTCs) from blood through a seven-to-10 day process that begins with the removal of red blood cells and certain white blood cells via antibodies and a magnetic field, followed by fluorescence in situ hybridization (FISH) with probes targeting DNA that is amplified in lung cancer, fluorescence microscopy, computer-assisted image sorting to identify likely true CTCs, and a final human review of the results.
While LungLife AI currently uses commercially available image analysis and sorting software, it plans to include its own machine learning (ML)-assisted image analysis in the future to reduce both technician labor and the risk of misclassification.
The company plans to file for marketing approval from the US Food and Drug Administration and has initiated a pivotal clinical validation study of LungLB in up to 425 participants. The study is currently recruiting, with enrollment expected to be completed in April or May and data analysis in July or August. The primary endpoints of the study will be assay sensitivity and specificity.
This study follows a recently completed pilot study of LungLB's efficacy in 149 patients with indeterminate nodules. Participants were tested by LungLB and subsequently biopsied to confirm disease status.
According to LungLife CEO Paul Pagano, testing assay performance in patients with indeterminate lung nodules was critical, as this is the patient population that is most challenging to diagnose.
"It's that intermediate risk group where the nodule is a little bit on the smaller side [and] maybe they don't have a strong smoking history, where biomarkers are most valuable," he said.
The pilot study results, which are currently under review for publication in a journal, indicated that LungLB has a 77 percent sensitivity, detecting 85 of 111 malignant nodules, and 72 percent specificity, correctly identifying 27 of 38 benign nodules.
Pagano mentioned that the main change in study design between the pilot and pivotal studies will be that the nodules examined in the pivotal study will be limited to those no greater than three centimeters in diameter, making them smaller than the ones assayed in the pilot study.
"That's kind of our sweet spot," Pagano said. "The smaller nodules are potentially earlier-stage cancers, are harder to test, and that's where we work best."
For comparison, Veracyte's Percepta Nasal Swab assay measures gene expression signals associated with cancer-related tissue damage in the lungs. That test showed 97 percent sensitivity and 40 percent specificity among patients considered to have low risk for lung cancer, and 92 percent specificity and 57 percent sensitivity among patients deemed high risk. Veracyte noted that sensitivity is the more important metric in the low-risk group so as not to miss the presence of cancer, while specificity is more important in the high risk group for avoiding false positives.
Biodesix, another competitor in the lung cancer nodule space, uses a combination of two proteomic liquid biopsy assays to risk-stratify patients, where one test — Nodify XL2 — is optimized for high sensitivity and low specificity and the other — Nodify CDT — for high specificity and low sensitivity.
Goodbody, an investment bank that recently initiated coverage of LungLife AI, wrote in a recent analyst report that although Nodify appears effective in assessing lung cancer risk, the use of two separate tests increases both complexity and price.
Nodify XL2 has a list price of $3,850, for instance, while in November of last year, the Centers for Medicare & Medicaid Services handed LungLife AI a final pricing determination of $2,030 for LungLB for each eligible Medicare patient.
On the technological side, LungLife's magnetophoretic CTC enrichment platform makes it one of the few companies to offer such devices commercially.
Other companies in that space include South Korea's GenoBio, Italy's Menarini Silicon Biosystems, and University of Georgia startup FCS Technology.
LungLife AI distinguishes itself from these other companies largely by isolating CTCs through "negative selection," in which antibodies are directed against other, non-tumor cells. Antibody and magnetic bead-labeled cells are then separated from unlabeled cells by a magnetic field.
Pagano explained that LungLife AI chose negative selection because of the ways that CTC surface protein expression changes over time, whereas membrane proteins on healthy blood cells tend to be more stable.
"We didn't want to use positive selection and fear that we might miss some of those cells that downregulate their markers when they enter the bloodstream," he said. "Knowing what we don't want to look at and knowing that those markers are very well-defined and stable allowed us to have a very robust assay."
In contrast, both GenoBio and Menarini use positive selection strategies, labeling CTCs with antibodies such as EpCAM, vimentin, and c-MET, whereas FCS Technology separates cells by size.
Although CTCs do tend to be a little larger than other cells, Pagano said that "these cells can range in size, and the range is fairly large. We've found some [CTCs] that are on the order of 10 microns, all the way up to a cell that's 30 or 40 microns in size."
GenoBio recently launched an updated version of its GenoCTC platform — GenoCTC v5 — in South Korea and anticipates an international launch by June of this year.
The GenoCTC v5 is for research use only, and the updates include improved separation, an integrated operating system to improve user convenience, and a reduction in size and weight by over 50 percent. CTCs isolated on the GenoCTC v5 platform are then analyzed with Bioview's DeNovo imaging and analysis instrument.
At the recent Tri-Con summit in San Diego, Chaithanya Chelakkot, scientific adviser to GenoBio, presented data from two clinical studies, which the company intends to publish in peer-reviewed journals this year. The company is also currently testing the GenoCTC v5 in several international clinical studies aimed at evaluating various aspects of EpCAM-positive and c-Met positive CTCs in cancer settings.
Similar to GenoBio and FCS Technology, Menarini's CellMag platform is also for research use only. At present, the CellMag kit is only for epithelial CTCs, but the company is exploring kits for the enrichment of other rare cell types. It is also testing the possibility of adapting its CellMag workflow to perform parallel EpCAM-dependent enrichment of tumor-derived extracellular vesicles and CTCs from the same blood sample.
Menarini will present preliminary data on this effort late this month at the American Association for Cancer Research conference.
LungLife AI aims to file for regulatory approval of LungLB via the de novo pathway, which is designed for low- to moderate-risk devices that, due to their innovative nature, lack competing products that are similar enough for direct comparison.
It joins Angle, manufacturer of the Parsortix cell isolation platform, in seeking marketing approval for a diagnostic CTC platform. While LungLife AI targets lung cancer, Angle plans to file for approval of an ovarian cancer test in the near future.
Although LungLB is currently only being tested for use in evaluating indeterminate lung nodules, it can also be applied to post-surgical disease monitoring and could in theory be used as a prognostic measure in predicting the risk of disease recurrence.
Pagano explained that even with early detection, there is an approximately 50 percent chance of recurrence in lung cancer, which frequently appears as metastatic disease.
"What that suggests," he said, "is that circulating tumor cells are present at the time of diagnosis, but they're not detectable by CT scan, so [the cancer] may have already spread to different parts of the lung, liver, or other organs. But it might be as a single cell or as very small cellular clusters called micrometastases."
LungLife AI currently offers LungLB to physicians through an early-access program.
Through this program and its planned clinical validation and utility studies, the company will be collecting information on whether and how LungLB impacts clinical decision making and helps to avoid unnecessary invasive procedures, as well as the device's overall benefits to patients and to the healthcare system.
"These are things that payors will look for when making a coverage decision," Pagano said.
LungLife AI is currently talking with other public and private payors about potential coverage and pricing agreements, but it remains too early for any to be finalized.
The company recently obtained both CLIA and CLEP certifications for its laboratory, enabling it to commercialize LungLB in 47 US states, which it expects to begin by the end of the year.