I was extremely excited for this afternoon’s New Horizons lecture on early detection of cancer by Sanjiv Gambhir, MD, PhD, and his talk did not disappoint. We naturally tend to focus on our existing methods of cancer diagnosis and treatment, forgetting that CT wasn’t a commonplace technology until the 80s, and that PET and PET/CT are even more recent developments. Gambhir’s overview of Stanford’s plan for more effective early detection of cancer was, therefore, inspirational and exciting. What innovation will be the PET/CT of the next decade?
Gambhir began with a discussion of the limitations of cancer treatment today. Although many of our treatments are revolutionary in their execution, they’re not as effective as they might be because late-stage cancer remains difficult to eradicate. Survival as a function of stage is identical for almost all cancers, Gambhir noted, and as a result we wind up investing nearly 100 times more in late-stage cancer treatment than early-stage. “This has to change – at the research level and at the level of bringing in private money to commercialize diagnostic strategies,” he said.
According to Gambhir, the future of cancer treatment will rely on what he calls “the three Is”: identification (through a low-cost test such as a blood test), imaging and intervention. Blood tests will serve as a “front gate” to determine who should go on to receive imaging. At Stanford, the school of medicine has collaborated with the school of engineering to create an incredibly sensitive biochip capable of detecting the proteins shed by cancers even at an early stage.
Moreover, researchers have shown that applying a pulse of ultrasound to the region of the suspected tumor – or, as Gambhir humorously put it, “yelling at the tumor” – can cause it to shed extra biomarker proteins, rendering blood tests more definitive.
After catching the cancer earlier than ever before – in one trial of Stanford’s magnetonanosensor (they’ll need a catchier name!), blood tests were able to detect ovarian cancer an average of 18 months before clinical diagnosis – we’ll still need imaging. But Gambhir is seeking a new kind of imaging technology, one that fits all of his dream criteria: anatomically informative, molecularly informative, good spatial resolution, intrinsic molecular signal, extrinsic molecular signal, multiplexing capability, high sensitivity, high throughput, safe, and – here’s the kicker – low-cost. “It’s a very tall order indeed,” he said, “but if we don’t set the bar high, we’ll never get to realizing a solution.”
Some emerging technologies cited by Gambhir that begin to meet these criteria include targeted microbubble ultrasound (entering into clinical trials in 2011), molecular endoscopic imaging and photoacoustic molecular imaging, in which red-shifted light is used to heat tissue, leading to thermoexpansion that creates an acoustic detection. Here’s a shot of the first-ever human breast image attained this way – as you can see, it’s incredibly high-res and deep in its penetration:
“This is just the beginning,” Gambhir concluded. “These technologies will fundamentally change our ability to see events at an earlier level, allowing the success of what are already good interventions, but are being used too late.”