Finding cancer earlier means patients have a better chance of being treated successfully. At AACR, a panel of researchers came together to present their various strategies for earlier detection of various cancers. Stanford's Sanjiv Gambhir said one problem is that the scientific community invests more in late-stage testing and treatment than in early detection. He added that because not all tumors kill patients, early detection must be balanced with taking care that clinicians don't over-diagnose. The key to early detection is in finding more biomarkers of diagnosis and prognosis, and this will come to fruition through both in vivo and in vitro diagnostics. Gambhir and his team are working on new ways to make cancer tumors shed biomarkers as early as possible without provoking the tumor into developing or metastasizing. Several researchers are experimenting with ultrasound waves while others are experimenting with different kinds of imaging technologies — like targeted microbubble imaging and photoacoustic molecular imaging — that are both more sensitive and less toxic than imaging technologies already in use. Many of these techniques are still only applicable to targeted body parts and not the whole body, Gambhir said, but they are in development and researchers hope to translate them to the clinic.
UCSF's Thea Tlsty took the podium to talk about cancer detection methods that aim to find cancer so early that it hasn't even formed yet. In various trials done during the last decade, researchers have found that CD36 — which regulates adipocyte differentiation — is under-expressed in invasive breast cancer tissue as compared to normal tissue. Since it is also know that high-density mammographic breast tissue is a risk factor for cancer, Tlsty and her team wondered whether CD36 was expressed differently in high-density or low-density mammographic tissue. They found that although low-density cancerous tissue had under-expressed CD36 levels, those levels were even lower in high-density tissue. This means, Tlsty said, that mammographic density of breast tissue and CD36 levels could be used as phenotypes for risk, potentially in the clinic, and that knowing these markers could lead to non-invasive imaging techniques and novel therapeutics for breast cancer, or any cancer that under-expresses CD36.
Also on the panel was Samir Hanash of the Fred Hutchinson Cancer Research Center, who spoke about the use of mouse models to find circulating protein biomarkers for cancer. Mouse models can now "recapitulate human cancer fairly well," Hanash said. He and his team developed a strategy to find low-abundance proteins in mouse blood to see which are under-expressed or over-expressed in control mice versus mice with cancer. Using this method, Hanash identified 54 potential markers of pancreatic cancer, which are being validated in human trials. Working with mouse models of human pancreatic, lung, breast, and other cancers, he and his team have developed a database of more than 7,000 protein biomarkers that could at some point be of help in the clinic for earlier detection of cancer.