NEW YORK – Optellum and Volpara Health teamed up this week to market artificial intelligence (AI)-driven cancer detection tools to hospitals in the US, pairing Optellum's lung cancer capabilities with Volpara's suite of tools for analyzing breast imaging.
Oxford, England-based Optellum develops AI-driven solutions for lung cancer. Its Virtual Nodule Clinic analyzes imaging data to determine the likelihood that a lung nodule is cancerous. The platform comprises a natural language processing algorithm and a deep learning algorithm trained on CT images from the National Lung Screening Trial. It identifies suspicious lung nodules found on CT scans and gives a probability score for cancer between 1 and 10, with 10 conferring the highest probability that a suspicious lung nodule is cancerous. The analysis is typically conducted on scans that are ordered for reasons unrelated to cancer concerns, which Optellum says allows doctors to diagnose cancer and initiate treatment at an earlier stage.
The US Food and Drug Administration granted 510(k) clearance to Optellum's Virtual Nodule Clinic in 2021, and in September 2022 the company raised $14 million in Series A financing to advance the technology. According to Optellum CEO Jonathan Watkins, the technology is already in use at several healthcare centers in the US, including at Vanderbilt Health in Tennessee and Augusta Health in Virginia.
Now, by joining forces with Volpara, Optellum can expand access to Virtual Node Clinic to more hospitals, health systems, and imaging centers across the country and enable opportunistic screening and precision diagnosis of lung cancer.
The Virtual Node Clinic adds to Volpara's existing AI tool offerings, including a breast risk assessment platform that analyzes breast density, image quality, and compression pressure from mammography images for breast cancer screening and early detection. In May 2024, the South Korean digital pathology company Lunit acquired Volpara, headquartered in Wellington, New Zealand. Lunit has inked a number of deals around its AI-driven digital pathology and precision medicine technology, including a partnership with AstraZeneca signed in November 2024 to develop biomarker detection tools for lung cancer.
The latest partnership is primarily focused on the US market, but Watkins said Optellum may also collaborate with Volpara to reach international markets. Currently, the business partnership between Optellum and Volpara doesn't involve technical integration of the two platforms. "We will go together to providers, hospitals, and health systems with a combined offering on breast cancer and also on lung cancer, but no software integration at the moment," Watkins said. The companies have not disclosed the financial terms of their agreement.
Watkins sees the value of Optellum's technology for lung patients as an opportunity to "stage shift" the diagnosis from late to early stage. "At present a lot of lung cancers get diagnosed at a very late stage," Watkins said, noting that the five-year survival rate for metastatic lung cancers is a bit less than 10 percent. With Optellum's solution, the company is hoping to enable earlier diagnosis, he said, and is shooting to raise the five-year survival rate in metastatic lung cancer patients to above 90 percent.
In practice, when patients present with incidental lung nodules from a CT scan, the pulmonologist or thoracic surgeon uses the Virtual Lung Nodule Clinic to risk stratify and prioritize those at highest risk for cancer. "All of that enables a more streamlined diagnostic workflow and faster times to diagnosis," Watkins said.
In a pooled retrospective analysis published in Lung in May 2024, Optellum's technology was 11.6 percent more sensitive than a physician's judgment alone for detecting cancerous lung nodules and 14.5 percent more sensitive than clinical risk models.
While the current platform is designed for diagnostic purposes only, Optellum is pursuing development of its technology for treatment optimization and as a companion diagnostic. At the American Society of Clinical Oncology's annual meeting in May 2024, Optellum researchers presented results from a study that explored the accuracy of the machine learning model in predicting lung cancer recurrence following curative surgery. Among a group of 323 lung cancer patients, it predicted recurrence preoperatively with 35.3 percent sensitivity and post-operatively with 26.5 percent sensitivity. The model, the authors concluded, outperformed traditional methods of predicting recurrence based on lung cancer staging.
"As a result [of AI-based predictions], you can optimize the treatment choice by surgeons or oncologists," Watkins said, adding that the company is working with a number of partners to advance those capabilities.