Update: Zolt谩n Tak谩ts has now joined Jeremy Nicholson鈥檚 team at Imperial College London. Working together, the group have tested the smart knife in action 81 times on people with cancer. The tool correctly identified both healthy and cancerous tissue, matching the post-operative histological diagnosis in all trials ().
Original article, published 2 February 2011
THE smell of burnt flesh rises in the operating theatre and the smoke from vaporised tissue is sucked away. But these fumes are not channelled out through the ventilation system. Instead, they are diverted into a machine that tells the surgeon exactly what is being cut into, guiding the rest of the operation. This is 鈥渟mart surgery鈥�, and it holds the potential to transform medicine.
Advertisement
Timely tissue analysis is also high on the agenda outside of the operating theatre. When confronted with an ambiguous lump during a hospital appointment, a doctor鈥檚 normal port of call is the histologist 鈥� a laboratory scientist who identifies tissue samples using a microscope. The process tends to take about 40 minutes, and is subject to human error and variability. To standardise and speed up tissue identification, and his colleagues at Imperial College London have brought nuclear magnetic resonance spectroscopy 鈥� a chemistry-lab staple 鈥� into St Mary鈥檚 Hospital in London. 鈥淭his is the first NMR spectrometer in a hospital in the world,鈥� says Nicholson. The idea is that the machine can provide an accurate read-out of the molecular make-up of a pinhead-sized tissue sample in mere minutes.
鈥淭he machine provides an accurate read-out of the molecules in a pinhead-sized tissue sample鈥�
NMR spectroscopy is most commonly used in chemistry to understand protein structures and chemical reactions. It utilises a property of the nuclei of different molecules that causes them to behave differently in a magnetic field: each absorbs radio waves at characteristic frequencies, enabling the machine to identify all the molecules in a sample.
鈥淏y using it in a hospital, we鈥檙e looking at [a diagnosis] in about 10 minutes,鈥� says , a surgeon involved in the project.
The technique could also cut down on costs, says Kinross. 鈥淭he machine costs roughly around 拢200,000, but the cost of each analysis is very low,鈥� he says, 鈥渁nd histologists are very expensive.鈥�
The next step is to take the technique into surgical practice. Figuring out whether a lump is benign or a cancerous tumour, or where cancerous tissue stops or starts, can be a tricky business for a surgeon at the operating table. 鈥淎n inflammatory mass can look and feel the same as a cancer,鈥� says Kinross. 鈥淚n those cases you perform the safest operation for the patient, and that might mean an extensive resection.鈥�
If a surgeon needs to identify a particular tissue, the operation is put on hold for the 40 minutes it takes a histologist to analyse a sample. 鈥淭he surgeons just stand around and tap their feet while the patient is left open on the operating table,鈥� says Kinross.
But what if surgeons could immediately work out what type of tissue they are about to dissect? at the University of Giessen in Germany have adapted another kind of spectrometer in Debrecen, Hungary, to do just that.
Surgeons rarely use a simple steel blade in theatre. Instead, a range of cutting tools are used which generate heat to both cut tissue and reseal blood vessels. A common method is electrosurgery, whereby a person lies on a giant flat electrode, and the surgeon cuts into them using a sharp electrode shaped like a knife (pictured). The electric current between the two electrodes produces heat at the point where the knife鈥檚 tip touches the tissue.
鈥淒uring the surgery there is smoke and a smell of burnt flesh,鈥� says Tak谩ts. This smoke is normally pumped out through a ventilation system, but his group redirects the fumes into a mass spectrometer (see diagram).
As the smoke is sucked up, the small molecules are separated from the larger pieces of tar, made from burnt flesh, by virtue of their different speeds through the tube. The small molecules then enter a mass spectrometer, which Nicholson describes as 鈥渁n extremely accurate weighing machine for molecules鈥�.
When tissue is vaporised, its individual molecules become charged. In the spectrometer, the charged molecules are accelerated in a vacuum by magnets. The larger a molecule is, the slower it travels, so the molecules are weighed 鈥� and identified 鈥� by their arrival time at the analyser at the end of the machine.
The information collected 鈥� the sample鈥檚 spectrum 鈥� can then be matched against a database of spectra of other tissue samples to pin down the tissue type, which is presented on a computer screen. Tak谩ts鈥檚 growing database contains about 16,000 spectra, which can identify a range of healthy and cancerous tissues.
All of this takes a mere 0.9 seconds, but by improving the algorithm used to check spectra against the database, the team predicts it can lower this even further, to about 0.2 seconds 鈥� effectively instantaneous.
Tak谩ts鈥檚 group are part-way through a clinical trial comparing patient outcomes in mass spectrometer-aided surgery with normal procedures. 鈥淲e haven鈥檛 yet compared them systematically but one thing鈥檚 for sure 鈥� this mass spectrometer-based tissue identification surgery decreases the surgery duration and invasiveness,鈥� says Tak谩ts.
The group is now looking for spectra that identify tissue surrounding a tumour. In most cases surgeons want to avoid cutting directly into a tumour, which could fragment and spread through the body. Ideally, surgeons would have a warning signal when cutting close to the cancer, enabling them to cut around it, says Tak谩ts, who hopes to find markers of inflammation that signpost nearby tumour territory.
The opposite is true for brain surgery. 鈥淗ere, the surgeons need to be alerted when the healthy brain tissue is being cut,鈥� says Tak谩ts, who has just begun trialling his technique in people with brain cancer. 鈥淭hey cannot cut around the tumour as they would be cutting into functional brain tissue. Millimetres count.鈥�
The technology could also be used to enhance surgical robots. 鈥淲hat we鈥檙e looking at is biochemically aware robots,鈥� says Kinross. 鈥淚 can鈥檛 think why you鈥檇 spend all that time and money to build a robot and not integrate into it a sensing tool which improves decision-making.鈥�
Both research groups are hopeful of transforming surgery and hospital treatment. , an analytical chemist at Purdue University in West Lafayette, Indiana, says he is cautiously optimistic. 鈥淭ak谩ts has made an important advance in applying this technology to surgical smoke,鈥� he says. 鈥淚t鈥檚 a clever way to go.鈥�
