Drug attrition is one of the defining problems of the pharma industry. Of the many promising drug candidates that start Phase I trials each year, less than 10% will ever reach marketing approval. The rest will fail during clinical trials, commonly due to problems with safety or efficacy.
Given these low success rates, it is hardly surprising that the costs of drug development are so high. According to the Tufts Center for the Study of Drug Development, it costs nearly $2.6bn, on average, to develop a new drug (including expenses of $1.4bn and $1.2bn in lost investor returns). If dead-end candidates could be eliminated early on, this price tag would drop significantly.
This, at any rate, is the thinking behind the 3D OrbiSIMS, a new molecular imaging instrument developed by the UK’s National Measurement Institute (NPL). As reported in Nature Methods in November 2017, this device can image drug uptake at the cellular level, helping identify the kinds of anomalies that lead to late-stage failure.
Developing measurement capabilities for subcellular imaging
“Around five years ago, colleagues from GSK came to visit NPL and discussed their challenges in this area of drug attrition,” recalls Professor Ian Gilmore, senior NPL fellow and founder of the National Centre of Excellence in Mass Spectrometry Imaging (NiCE-MSI).
“They wanted to see if we could spot where drugs go at a cellular scale, to find out whether the concentration is at the right kind of therapeutic level. They also wanted to find out whether the drug is going into certain organelles, which can be an indication of toxicity,” he adds.
It wasn’t an easy ask. The average cell is around 20 microns across, with organelles such as mitochondria measuring just a micron. At the time of GSK’s visit, the existing instrumentation wasn’t capable of detecting drugs at a subcellular resolution.
“It wasn’t possible with the techniques we had to get down to the resolutions they wanted,” says Gilmore. “But one of the roles of a national laboratory like ourselves is to innovate new measurement capabilities, and I had an idea around that time for a new instrument we could develop.”
Speed versus sensitivity: getting the best of both
The problem with existing mass spectrometry devices was that they always entailed a trade-off between speed and sensitivity. On one hand, there were super-speedy time of flight (TOF) analysers, which weren’t very precise. On the other hand, there were instruments such as the Orbitrap, which were accurate but slow. Gilmore’s idea was simple: why not combine two different types of mass spectrometers, and in doing so get the best of both worlds?
“The speediness of the TOF offsets the slowness of the Orbitrap, and the accuracy of the Orbitrap offsets the inaccuracy of the TOF,” says Gilmore. “I proposed the idea to GSK and they really liked the idea, so we formed a consortium to meet the challenge.”
This multidisciplinary team included drug discovery experts from GSK and scientists from the University of Nottingham and NPL. It also included two leading spectrometry companies, ION-TOF GmbH (a specialist in TOF-SIMS devices) and Thermo Fisher Scientific (manufacturer of the Orbitrap). Together, they integrated these technologies onto a single platform, the 3D OrbiSIMS.
Secondary ion mass spectrometry: how it works
In simple terms, secondary ion mass spectrometry (SIMS) works by scanning a focused ion beam over a surface. Every time it hits a pixel, molecules from the surface ‘sputter’ into the mass spectrometer, generating a mass spectrum and eventually a 2D image. 3D images are formed by removing layers from the surface, slice by slice.
In a typical TOF-SIMS instrument, the material that enters the device is discarded. With the OrbiSIMS, however, the molecules are trapped and analysed by the Orbitrap, meaning they can be identified with greater precision.
“The OrbiSIMS also brought to bear a recent development in the SIMS field, namely large gas cluster ion beams, which are much gentler than the beams we used in the past,” says Gilmore. “This gives better sensitivity in looking for molecules. As we demonstrated in our Nature Methods paper, we can look at a single cell and see the drug uptake there, but also measure metabolic changes, which is really important for this question of drug attrition.”
He cites Pfizer’s ‘three pillars of survival’, a key principle in drug development. If a drug is to evince any therapeutic benefit, it needs to reach the site of action, bind with the target, and elicit metabolic changes. Uniquely for an imaging device, the OrbiSIMS will be able to image all three.
Gaining new insights into cell variability
Since the device is so precise, users will be able to see differences between one cell and the next (even those within the same cell line). These differences, imperceptible with earlier imaging techniques, may prove all-important when it comes to their interaction with a drug.
“In our paper, the cells were notionally all the same, but the amount of drug they took up would vary by an order of magnitude,” says Gilmore. “So for the first time, you’re starting to get insights into the cellular heterogeneity of drug uptake, which becomes a very powerful tool for the industry.”
The OrbiSIMS is also being used within the Cancer Research UK Grand Challenge project, which aims to create a ‘Google Earth view’ of cancer.
“It’s known that a tumour is not one simple cell type – there are lots of different regions, and our goal in this project is to be able to map these regions at a sub-cellular resolution,” says Gilmore.
Their next step, Gilmore says, is to make the instrument cryo-capable – in other words, suitable for use with frozen cells. Since cell metabolites change constantly, frozen biopsy samples will enable researchers to preserve the cells’ structural integrity and provide a snapshot in time.
Over the years ahead, they want to increase the resolution further, providing metabolic imaging below the diffraction limit of light.
“Super-resolution microscopy has absolutely transformed the field of cellular biology, because you can image all different types of protein while the cells are alive using a fluorescent probe,” says Gilmore. “You can’t use that fluorescent probe strategy when you want to look at metabolites or small drug molecules, so our long-term goal is to reach this super-resolution barrier without fluorescent labels.”
This would mean achieving a resolution below 250 nanometers, six times sharper than the current limit of 1.5 microns. It would provide a whole extra level of detail into cells and their metabolites.
Although that is some way down the line, the 3D OrbiSIMS is already being used within the pharmaceutical industry with a view to reducing drug attrition rates.
“If you can better understand where a drug’s going, whether it’s binding to the right targets, and whether it’s having the right pharmacological effect, then you have the chance to see earlier in that pipeline whether a drug is potentially going to fail. That stops a lot of wasted investment,” Gilmore explains.