Made to measure: MRI technique to assess lung cancer16 February 2018
A new MRI technique might provide a reliable way to assess how lung cancers are responding to treatment. Kerry Taylor-Smith speaks to study leader Professor Nandita deSouza, translational imaging professor at the Institute of Cancer Research, about the tumour-measuring technology.
Average adult human bodies are around 57–60% water, while young babies are at least threequarters liquid. Most of it is present in intracellular fluid, which performs such functional duties as metabolising proteins and carbohydrates, and insulating and protecting vital organs. The water can also play a vital role in imaging.
Water consists of two hydrogen atoms and one oxygen atom. The proton at the centre of the former can be driven to produce the signals that generate MRI pictures. The technique uses strong magnetic fields and radio waves to produce detailed internal images of the body. The protons, which are sensitive to magnetic fields, act like tiny magnets. When a patient is inside an MRI scanner, all the protons align along the field, a bit like a magnet pulling a compass needle. The protons are then knocked out of alignment by short bursts of radio waves, and realign when the waves stop, sending out signals that are picked up by receivers.
The position of these signals can be precisely located and their nature helps distinguish between the various types of tissues in the body, because protons in different tissues realign at different speeds and produce distinct signals. Just as millions of pixels on computer screens join together to create complex pictures, the signals from the millions of protons in the body are combined to create a detailed image of the inside of the body.
This makes the technique particularly useful for measuring tumours. While MRI can examine the brain and spinal cord, bones and joints, and the heart and blood vessels easily, the lungs have been difficult to image because the proton-poor air they contain creates weak magnetic fields that can distort images.
PET scans use a nuclear medicine imaging technique that produces images of functional processes in the body. They require small amounts of a radioactive drug – typically the radiopharmaceutical fluorodeoxyglucose (FDG) – to show differences between healthy and diseased tissue. A small amount of FDG is injected into the patient and, because cancers grow quicker than healthy tissue and metabolise more glucose, their cells take up more of the radioactive FDG. A PET scanner detects radiation given off by the FDG, and this signal is translated to a colour map of the body that shows areas of high glucose uptake, indicating the presence of cancer.
“When cancer cells respond to treatment, tumour cells may die, but the shrinkage lags behind,” says Nandita deSouza, translational imaging professor at the Institute of Cancer Research (ICR). “Metabolic response on PET is a better indicator of response but, particularly with new treatments that target the immune system, the immune response to these therapies, which results in increased glucose uptake, means that the metabolic response is not a reliable marker.”
In the new study, scientists at the ICR and The Royal Marsden NHS Foundation Trust took scans, twice over, from 23 patients at four different hospitals. They measured the effect of using different types of software to process the data produced during the scan – a critically important point if running a treatment trial in multiple centres, because different hospitals have access to different technology.
The researchers recorded the movement of water within tumours during the MRI scans using ADC, which measures the magnitude of diffusion of water within tissues, and is calculated from MRI that uses diffusion-weighted radio wave pulses. The technique sensitises the signal detected to the random motion of water molecules. It is a measure of the extent of tissue cellularity, because the presence of intact cell membranes impedes water molecule diffusion.
“Diffusion-weighted imaging generates tissue contrast from the movement of water within tissues,” says deSouza. “On heavily ‘diffusion-weighted’ sequences, which pick up signals from protons moving short distances, the signal from protons in tissues where cells are loosely packed diffuses away and very little signal is left. In comparison, signals from cancer tissues are retained, because the cells are tightly packed and the distance water moves is restricted by cell membrane boundaries.
“ADC measures this rate of signal decay as more diffusion weighting is applied,” deSouza continues. “ADC is low in cancers where the signal is retained and does not diffuse away, despite increasing diffusion weighting. Conversely, it is high in loosely packed issues and obviously highest in free water. ADC provides a quantitative measure of tissue cellularity, which is related to how cellular, and thus how aggressive, the tumours are.”
The study, which was published in the journal European Radiology, found that measurements varied by less than 10% between scans across all patients tested and were consistent across hospitals, meaning measurements were not affected by where the scans were taken, or the equipment and software used during the process.
Importantly, the patients could breathe freely during the scans, which is often not possible during scans for routine care at the moment. “It can be difficult to assess lung cancers using MRI, firstly because the patient must continue to breathe during the process, so the chest is moving, and because the air in the lungs can cause problems with the readings,” explains deSouza.
There was less than 4% difference between ADC values for tumours larger than 3cm in diameter. Measurements of smaller tumours, which can be more affected by the movement of breathing, still showed a high degree of repeatability, with a maximum variation of less than 10% between scans.
“Measuring the apparent diffusion coefficient of lung cancer with MRI is convenient and non-invasive, but this is the first time research has shown ADC values can be reliably measured using repeat scans in a multicentre setting,” says deSouza. “Our study shows that measuring changes in ADC for lung cancer is manageable for patients, with comparable results using different MRI. Importantly, it proves that we can proceed with multicentre lung cancer trials despite differences in the technology between hospitals.”
Although the sample group was fairly small, deSouza predicts that the measurement is useful and will be employed in lung-cancer treatment trials.
“Remarkably, though, it is a robust measurement with less than 10% variation on repeat scanning; so using it to monitor treatment, where a much bigger change than that is expected, means that it can be used relatively reliably,” she concludes.
The study will pave the way for clinical trials of the technique in lung cancer and could enable doctors to take more accurate scans than is currently possible. The technique has already been used on many other cancer types, and lung cancer is one of the last diseases in which it has been employed. Despite the challenges, deSouza believes the technique could shed vital new light on how lung cancers respond to treatment.