'Did you get it all?' Experimental 'pen' could help surgeons detect remaining cancer instantly

A prototype "pen" technology aims to give cancer surgeons a better sense of whether they've removed all of a tumour while in the operating room.

In samples taken from 253 people in U.S., device was 96% accurate in identifying cancers, researchers find

Scientist Jialing Zhang demonstrates using the MasSpec Pen to analyze a human tissue sample. (University of Texas at Austin)

A prototype "pen" technology aims to give cancer surgeons a better sense of whether they've removed all of a tumour while in the operating room.

The new approach strives to answer a patient's question after surgeons remove a tumour: "Did you get it all?"

To say yes confidently, doctors often send a sample of the tissue in question to a pathologist to freeze, check and test under a microscope. It's the gold standard for accuracy, but takes at least 30 minutes per sample. As the clock ticks, the patient remains under anesthesia.

Cancer surgeons estimate that in up to 20 per cent of some operations, traces of tumour remain at the margin or edges of where they've cut, which can lead to repeated surgeries or recurrence of cancer.

In this week's issue of the journal Science Translational Medicine, U.S. researchers describe an experimental pen-like tool for surgeons that quickly draws samples through a tube to test for traces of cancerous tissue in the OR, with results within 10 seconds.

The technology could achieve the goals of more precise, quicker and safer surgery, said study co-author Dr. James Suliburk, chief of endocrine surgery at Baylor College of Medicine in Houston.

"It allows us in real time to test tissue from the operative field as we actually operate and ensure that we've resected the tumour to appropriate surgical margins," Suliburk said.

In a series of experiments, investigators in Texas tested the pen-sized prototype tip that's connected to a conventional mass spectrometer, which varies from the size of a laser printer to a fridge.

In validation tests on samples taken from 253 people with known lung, ovary, thyroid or breast tumours from a tissue bank, the device was 96 per cent accurate in identifying cancers, the team said. 

Giant leaps forward

The researchers also used what they call the "MasSpec Pen" to reliably identify breast tumours in mice. They hope to use it in surgery on human patients early in 2018 followed by clinical trials.

"Usually we take small step forwards and along the way we take several steps back," Suliburk said. "Because of our multidisciplinary research team … we've been able to take giant leaps forward with only small steps backwards. And that's something that in modern-day medical research seldom happens."

Suliburk envisions the technology eliminating the need to send out tissue to be analyzed under the microscope. 

Researchers say the MasSpec Pen can non-destructively analyze human tissue samples to identify cancer. (University of Texas at Austin)

Livia Schiavinato Eberlin, an assistant professor of chemistry at the University of Texas at Austin, envisioned the technology. Co-author and campus colleague Thomas Milner, a professor of biomedical engineering, worked with her team to put together a prototype that could be held against a patient's tissue.

It's biocompatible, since it just takes water deposits from the surface of the tissue. Then samples are transported through special tubing to the mass spectrometer, said Milner.

Deciphering tumour fingerprints

Next, machine learning software analyzes the "fingerprint" of metabolites from the sample to find predictably higher concentrations within a tumour. In seconds, the surgical team sees the final result on a screen displaying "cancer" or "normal." 

"Because the imaging and tools are so good, these treatments can be done predictably and reliably," Milner said.

There's a whole fleet of similar technologies coming to the operating room. For instance, the iKnife and its accompanying mass spectrometer samples the sharp-smelling smoke that rises when surgeons heat tissue as they cut. Other experimental modes are based on ultrasound or light.

Conventional pathology will remain paramount until it's proven that the molecular fingerprint from samples taken from next to a tumour are truly a cancerous cell, cautioned Dr. John Rudan, head of surgery at Queen's University in Kingston, Ont. He was not involved in the study.

"How does the clinician react to that area that happens to light up in the tissue adjacent to where you think the tumour is? We don't know what the answer is going to be yet. It's going to take a while for us to actually understand that."

Rudan sees how the MasSpec Pen could help surgeons to determine if surgical margins are clear. He's also aware of limitations on what it could accomplish. For instance, tumours are three-dimensional and need to be scanned all over, yet the tool itself lacks depth perception.

"I see this as a very interesting type of technology that will probably have a place in surgical oncology," said Rudan, who holds the university's Britton Smith Chair in surgery.

At Queen's, Rudan and his colleagues are using an image-guided tool on frozen breast cancer specimens. Part of the challenge is that breast tissue moves between the time when it's sampled, removed and analyzed. Without tracking technologies, surgeons can be working blind when they return to a particular spot.

The MasSpec Pen prototype hasn't been used in a real surgery. It could take years to determine if it's valid, cost-effective and ultimately improves survival rates.