UT researchers create pen that detects cancer in 10 seconds

Areeba Khwaja

Some pens write, others detect cancer.

UT researchers created a device called the MassSpec Pen, which can detect cancer during surgery within 10 seconds.

The device was envisioned by Livia Eberlin, UT assistant professor of chemistry, who said there was a need for it in surgical resection procedures where surgeons weren’t able to quickly determine the margin between cancerous and normal tissue.

Typically, when a patient undergoes surgery to remove cancerous tissue, the surgeon must first cut out the tissue and freeze it. According to research assistant Noah Giese, it is then sent to a pathologist, a scientist who studies diseases, who looks at it under a microscope, determines whether it is cancerous tissue or not and reports back to the surgeon.

“That entire process requires about 30 minutes or even an hour,” biomedical engineering professor Thomas Milner said. “The Pen does a measurement in 10 seconds.”

Oftentimes, the surgery may take so long that the patient is sewn up and asked to return the next day, according to Giese.

“We really tried to make this a product that can help patients and surgeons both, because current procedures are suboptimal,” Giese said. “While other procedures can irreversibly damage the tissue, the MassSpec Pen doesn’t do that — it’s completely non-invasive, biocompatible and safe.”

The Pen works by depositing a drop of water on the tissue surface, according to Giese. Biomolecules, or organic molecules in all living organisms, from the tissue will move into and interact with the water droplet.

“We used the most common solvent on Earth: water,” Giese said. “You can think of it as water absorbing the coffee beans when you make coffee, or absorbing the dirt from your body when you shower, it’s a simple diffusion process.”

The water is then analyzed by the mass spectrometer in the Pen which processes the ionized molecules and sorts them by their mass-to-charge ratio, according to Giese. Then, machine learning software analyzes the sample and determines whether the tissue is cancerous.

“Mass spectrometry is a very powerful technique that can detect lipids, metabolites and sugars,” Giese said. “We built our classifier machinery around it to predict whether a given sample would be cancerous or not by comparing the ratio of these biomolecules to normal tissue.

Giese added that the Pen is easy to use in real time.

“During surgery, it’s like a spot treatment,” Giese said. “Surgeons can use it optically by simply outlining multiple spots around what they believe to be cancerous tissue and then using the Pen to confirm or disavow their hypothesis.”

Milner said that the Pen can be trained for more specialized use in cancer detection.

“The molecular signatures for each type of cancer need to be identified by a machine learning software,” Milner said. “There is no limit on the types of cancer that can be detected — the system simply needs training.”

As of now, the Pen has been tested on four different cancers including ovarian, breast, thyroid and lung. The researchers are looking to test it on more types of cancer and even predict the subtype of a particular cancer.

“The development of the MassSpec Pen was something that I envisioned could really make a significant impact in cancer patient treatment, and I really believe that with further development it could be adopted in hospitals worldwide to improve cancer diagnosis and surgical treatments,” Eberlin said.

The team is now looking towards starting clinical trials with sponsoring institutions next year.

“It is so rewarding to … see this idea becoming a reality,” Eberlin said.