The Circle of Life

At a Glance

• Commonly used endogenous biomarker cancer tests often have low sensitivity and specificity and high background expression from normal tissue
• Exogenous biomarkers like engineered DNA minicircles have high sensitivity and specificity with little to no background expression
• The minicircles contain a tumor-activatable promoter and a reporter gene whose levels can be measured in the blood
• In the future, minicircles could also be engineered to contain therapeutic genes for cancer treatment

Most researchers looking for new methods of detecting cancer opt for endogenous biomarkers like proteins, nucleic acids or circulating tumor cells. There are obvious benefits to using a system like this – the markers are found in the bloodstream, enabling easy measuring; they offer affordable screening options; and measuring surrogate instead of true endpoints allows for smaller, faster trials with fewer ethical concerns (1). But the other side of that coin is the downsides of endogenous molecules: they often suffer from poor sensitivity and specificity because of low blood concentrations, rapid degradation, tumor heterogeneity and background expression in non-malignant tissues (2). In fact, it has been estimated that tumors may grow for as much as 12 years or longer (3) before current clinical biomarker assays can detect them, by which time they are highly likely to have metastasized. It’s plain that we need better strategies for cancer detection – but if not endogenous biomarkers, then what?

Building an exogenous biomarker

Sanjiv Sam Gambhir and his colleagues at Stanford University, USA believe they may have a solution to that problem. They’ve developed an “exogenous biomarker,” a DNA minicircle construct containing a tumor-specific promoter and a reporter gene that can be detected in the blood. In the case of this particular construct, the Survivin promoter (pSurv) is active in many cancers, but not in healthy adult tissues, so transcription of the human secreted embryonic alkaline phosphatase (SEAP) reporter gene is only driven in the presence of cancer. SEAP is a common reporter protein; it can be detected easily and with high sensitivity, has little to no background expression in adults, has low immunogenicity because it’s a human protein, and has already seen successful use in the clinic. Pairing pSurv with SEAP in tumor-activatable minicircles (see Figure 1) offers the chance to detect cancer in patients by simply administering a systemic dose of the minicircles, allowing time for gene expression, and then measuring SEAP levels – which should only be detectable in the blood of patients with tumors.

To test the tumor-activatable minicircles, the research team used mouse models. After definitively establishing that minicircles outperform traditional plasmids in melanoma cells, they injected their constructs intratumorally into mice with melanoma xenografts, which resulted in significantly increased plasma SEAP concentrations compared with control mice – and the results lasted for up to two weeks. Unfortunately, intratumoral injections aren’t always feasible in human patients, so their next step was to see whether a systemic injection of minicircles would have the same effects. Not only was the test able to discriminate easily between tumor-bearing and healthy mice, but the effects lasted well over a week, leaving a wide window of opportunity to read the test results. It was even possible to use minicircles to evaluate tumor burden. In mice with lung tumors, they measured the plasma SEAP concentrations at multiple time points and calculated the area under the curve; those values were closely correlated with the size of tumors in the lung, indicating that the new test can be used to assess not only the presence, but the extent of disease.

And the future?

Based on their mouse studies so far, the group believes the test looks very promising. Compared with endogenous biomarker tests, good sensitivity and specificity is evident – tumor-bearing mice can be distinguished from normal ones about 90 percent of the time. “In terms of sensitivity,” says John Ronald, lead author on the published paper, “I’d estimate that right now we can detect a tumor about the size of a grain of rice. Now that we can do that, we’re looking into making newer formulations so that we can detect smaller and smaller tumors.” He explains that one of the advantages of his group’s probes is that the strategy is very modular – they can test different delivery agents, different promoters, or different reporter genes. That lets them iteratively optimize the system to improve its sensitivity and specificity.

“I would say that we’re within five years of testing our first-generation construct in humans,” says Ronald. “Our laboratory has focused quite heavily on developing new invasive diagnostic technologies, and it generally takes us about five years to begin testing them in humans.” But such technologies do require a lot of safety testing – preclinical testing, regulatory issues, and many more precautions standard for clinical trials. The first step for the Stanford scientists is to test the minicircles in breast cancer patients with known tumors. If the minicircles can detect cancer in people with large tumors, then they can be tested on smaller ones to see how effective the test remains. Then, once it’s fully understood how the test works in those patients, it can be attempted in people who are being monitored for tumor recurrence, and then eventually in people who don’t have cancer, but are at risk – for instance, families with BRCA1 or BRCA2 mutations. Of course, the ultimate goal if the researchers can show safety and efficacy is to introduce this as a new blood test for screening the general population – but that’s a high bar to set so early in the game.

“My vision for this test is that it will be used to detect cancer earlier than our current methods can, or to confirm cancer diagnoses made by other tests,” says Ronald. “I don’t expect it to be used in isolation – there will always be new types of tests, but doctors could pair our test up with a less sensitive or specific one, like an endogenous biomarker test, to get a more complete picture. Imagine giving a patient a blood test with results that indicate a possibility of cancer; with our system, instead of ordering costly and time-consuming imaging, you can simply do another blood test to confirm the diagnosis.” Ultimately, he says, minicircles might be able to replace those early blood tests, but for right now, they are a tool that can be used to confirm diagnoses – which is a useful first step.

The road to better diagnostics

The Gambhir laboratory was inspired to create the system when they realized that the major problem with endogenous biomarker testing is the background expression of those biomarkers by normal tissue. One well-known example is the prostate-specific antigen (PSA) test for prostate cancer. The test fails to distinguish between people who have prostate cancer and those who just have an enlarged prostate, because normal prostate tissue also expresses that biomarker. “We asked ourselves, ‘Can we reverse the problem? Can we make the cancer express something that normal tissue won’t express, but that the cancer specifically expresses?’ We’re trying to flip the idea around by forcing the tumor to make something that’s not detected at all in healthy people.”

It’s an idea the team came up with about three years ago. “It took a little while,” Ronald reports, “because at the start, we were struggling to find the right vector. Injecting a gene-based vector is something of an unconventional approach to cancer detection, so we wanted to focus on using an appropriate vector technology. Minicircles are the minimum genetic element that is required to express a gene – and we also know that they don’t integrate when you inject them systemically, so that’s a good thing from a safety perspective.” He admits there have certainly been challenges along the way. “I think the biggest one was figuring out how to purify the minicircles enough to get the formulations with the in vivo transfection agent to work. It also took a little while to engineer the right formulations, and then to detect tumors in mice, but now that the system works, we can start to make better formulations.” That’s the step the researchers are working on now.

Ronald does acknowledge that it’s not quite all downhill from here yet. “I would say that, at the moment, there’s a lot of room for improvement in our next-generation constructs. We think they won’t just work for blood tests with reporter genes – minicircles can also be extended to therapeutic applications.” The group is currently also exploring the possibility of creating therapeutic constructs that express both reporter and therapeutic genes, so that they can be used for a combined diagnostic and therapeutic approach. “That’s the holy grail of cancer gene therapy – to express a therapeutic transgene specifically within a tumor so that healthy cells are not harmed. And it’s something we might have in the next few years!”

John Ronald is the lead author on the tumor-activatable DNA minicircle study and a postdoctoral scholar at the Molecular Imaging Program at Stanford School of Medicine, California, USA.