In the wide-ranging search for new weapons against cancer, some see great potential in the idea of a urine test that can reveal telltale signs of the disease. What this could look like and what role it might play in diagnosis and treatment are questions that are a long way from being answered, but in recent years we've seen a number of studies demonstrate what might be possible with these readily available samples of bodily fluid. So where does the technology stand right now?
There is so much going on in the human body that can affect the makeup of our urine, that pinpointing signs of cancer within it is a seriously difficult undertaking. Heightened levels of a certain protein might be indicative of the disease, but they could also be the byproduct of an entirely unrelated biological process, or perhaps because of something we ate or drank on that particular day.
"Urine is a difficult fluid to analyze, because there is so much content in there that could potentially be from any cell of the human body, so you need to know enough about all the different cancers and diseases for this to work," Dr Ian Johnson, research fellow specializing in cancer biology at the University of South Australia, tells New Atlas. "If you can use more than one biomarker, that increases your cancer specificity. Ideally you also need a baseline, to know what is “normal” in urine for an individual before they get a disease like cancer."
A more complete picture
If we know what a normal urine sample looks like for a particular patient, then it makes it possible to detect abnormalities that could spell trouble. But there is a lot to consider. While some proteins might go up in concentration over a particular timeframe, they might coincide with alterations in gene expression or other biological changes. For this reason, urine testing may prove most useful when it searches for a set of biomarkers, rather than just one.
A 2015 study from researchers in the UK offered a good example of what this approach could look like. Pancreatic cancer is very difficult to detect before it has spread and symptoms become apparent, and complicating the issue is that it is also hard to distinguish from an inflammatory condition called chronic pancreatitis.
The scientists, from the Queen Mary University of London, took aim at this problem by gathering 488 urine samples, including 192 from pancreatic cancer patients and 92 from patients with chronic pancreatitis. By analyzing these samples, the scientists identified a set of three proteins that all presented in higher concentrations in the cancer patients, and significantly lower levels in those with chronic pancreatitis. The team's hope was that this three-protein signature could one day form the basis of a urine test that detects pancreatic cancer in its early stages.
Spotting signs of trouble
As our knowledge of specific cancers continues to improve, it has downstream effects on the things we can look out for to detect them early in the piece. As a basic example, we now know a dash of blood in our urine could be a sign of bladder cancer, although once upon a time that wasn't necessarily the case. The same is true for scientists investigating the very fine biological details of the disease, and those developing tools to find signs of trouble.
"If you study enough of the same cancer in enough population, you can start to correlate what is released from the cancer that you can detect in urine, but of course it is mixed in with everything else that is going on in the body, so it is really difficult to pull that data apart and look for the most minute changes," says Johnson. "That's where new technologies are so important – to increase the resolution of detection enough that you can find just a tiny change in multiple different biomarkers."
Earlier this month, an interesting study from University of Cambridge scientists offered an exciting example of how this sort of thing might work. The research centered on cell-free DNA (cfDNA), which are fragments of mutated DNA that are released into the bloodstream by dying cells, including cancerous ones. In this case, the mutations seen on the circulating DNA fragments are the same as those in the primary tumor.
The Cambridge researchers started by identifying relevant DNA mutations on the primary tumors of brain cancer patients, enabling them to know what to look out for in a urine sample. A test subsequently designed to detect this cfDNA in urine did so successfully in 10 out of 16 samples. The team imagines this kind of tool could be used to screen recovering brain cancer patients at high-risk of recurrence, who might provide a urine sample every few months to potentially pick up the returning cancer cells at an early stage.
"We can detect the things that get packaged up and released from the cell, they might be bits of DNA, or mRNA, and proteins specific to the cancer origin," says Johnson, "But they’re in such low abundance in urine compared to everything else, that only now are we developing technologies that can pull these cancer-specific packages out to analyze in more detail and work out ways to detect them whilst they hide in plain sight.”
Another interesting example of this crossed our desk in June, when scientists in Japan demonstrated another urine test for brain cancer that boasted high accuracy. This one focused on microRNA, short-strands of genetic material that play an important role in gene expression, and appear in unique forms when produced by cancer cells in the body.
The researchers, from Nagoya University, developed a novel device with millions of zinc oxide nanowires that could extract vast amounts of microRNA from just a milliliter of urine. Samples were collected from brain cancer patients and non-cancer patients, with many of the microRNAs derived from brain tumors appearing on the device in stable condition. A diagnostic model was then developed and proved capable of distinguishing the brain cancer patients from the healthy controls with a sensitivity of 100 percent.
We're going to get to a point where we might need to detect biomarkers in urine constantly; every week, every month ...
While exciting, these kinds of examples are all couched in the idea that a physician would already know what the patient might be suffering from when taking the sample. However, it is possible that as the technology improves the scope of these tests could expand to less specific use cases, but would still likely be used as a way of confirming suspicions.
In this sense, any urine test that makes its way into clinical use in the near future is likely to be used as a technology that compliments other tools, rather than displaces them. And the need for a reliable baseline would be a further limitation, as recognizing irregularities that are representative of cancer with no background knowledge of the patient will likely be beyond the scope of the technology for some time yet.
"The key may be to consistently measure," says Johnson. "We're going to get to a point where we might need to detect biomarkers in urine constantly; every week, every month ... of course, you won’t go to the GP all the time, so there will be at-home sensors, perhaps in the toilet, that measure and record these levels, providing a baseline and record of biomarker changes that can highlight potential disease onset."
It does therefore seem that the idea of dropping into the doctor's office, providing a single urine sample and then waiting a few days for the all clear for cancer is off the table for now. It would require our knowledge of the many, many biomarkers to become so advanced, and for all the variables in every individual's physiological makeup to be accounted for, that such a reality is likely many years away. In saying that, it can't be ruled out entirely.
"Of course, never say anything is impossible," says Johnson. "If we can think it we can make it."
