Professor Kim Samkoe Lights the Way to Better Cancer Diagnosis and Surgical Precision

Unlike imaging techniques such as X-rays or radioactive PET scans, fluorescence is harmless and can be used as often and as much as a doctor wants without any ill effects on the body. Samkoe is collaborating with the Center for Surgical Innovation at Dartmouth-Hitchcock Medical Center (DHMC) to create colored maps of the structures of the body to aid in prostate cancer surgery—a kind of "GPS for surgeons," as she calls it. The effort is part of a $31.3-million contract across seven years from the Advanced Research Projects Agency for Health (ARPA-H), part of the "cancer moonshot," which aims to create novel surgical techniques to dramatically improve outcomes for the disease. In her lab's other work, Samkoe is also using her fluorescent superpowers to develop techniques to better diagnose and treat cancer with molecular therapies and customizing them to individual patients to maximize effectiveness.  

Color Maps for Surgery 

Growing up in Regina, Saskatchewan, Samkoe originally planned to become a doctor. At the University of Regina, however, she got sucked into a research project examining the structure of proteins, often the first step in developing targeted drugs. "I really loved the puzzle-solving aspect of it," she says, "knowing once you solve one problem, there is a next step and another—you are constantly figuring out new information." 

She first encountered fluorescence while studying for her PhD at the University of Calgary in Alberta, where she became involved in research to treat an eye disease in which blood vessels grow uncontrollably. Unlike bioluminescence, which produces light through chemical processes (think fireflies), or phosphorescence, which emits absorbed light slowly through time (such as glow-in-the-dark stickers), fluorescence works by exciting special molecules called fluorophores with visible light or other electromagnetic radiation. When hit by a shining light of one color, such as the ultraviolet "black lights" used at the dance club or bowling alley, the molecules emit light of a different color—meaning they can be activated at will and traced by a doctor or researcher.  

In the case of the eye disease, Samkoe and her colleagues were exciting fluorophores that can react with molecular oxygen and actually damage the blood vessels, stopping their uncontrolled growth. "This is where I developed my love of translational research"—the type of research that can transform discoveries in the lab into tools and therapies on the surgical bed. "Seeing the science you developed in the lab have 'real-world' benefits is highly rewarding." 

Coming to Dartmouth for a postdoctoral fellowship with Brian Pogue, now the Robert A. Pritzker Professor of Biomedical Engineering at Thayer, she changed her focus to applications in which fluorescence isn't being used as a therapy but to highlight cells and tissues to make other therapies and surgeries more effective. Along with Pogue, she developed techniques to attach fluorophores to cancerous tumors for the first time in humans, showing how they could highlight malignant tissues in head and neck, brain, and bone. "Kim was really running that project," says Pogue, who worked to create a professor position for her in the surgical department at Geisel School of Medicine. She later transferred to Thayer.  

In her current collaboration with the Center for Surgical Innovation around prostate cancer, Samkoe has expanded on those techniques by attaching fluorophores both to cancerous cells and critical normal structures such as nerves to better guide surgeons in removing tumors. Previously, surgeons just opened up the belly and cut out the entire prostate, a plum-sized organ surrounded by nerves and blood vessels, cutting through the urethra and nerves in the process, says Ryan Halter, a colleague and associate professor of engineering. "When you do that, a man has challenges after surgery with urinary incontinence and erectile dysfunction, so quality of life is really, really reduced," he says.  

Contemporary robotic surgery techniques offer less-invasive options, in which tools are inserted through small incisions and controlled remotely by doctors to remove tissue. Without being able to directly feel the tissue, however, they are often unable to differentiate tumors from other tissues. "It just looks like a red-and-pink mess," Halter says. Samkoe's innovations can light the path for surgeons to ensure removal of all the cancerous tissue while sparing nerves and blood vessels.  

Techniques she's developed even excite fluorophores with different wavelengths of light to tell exactly how deep the nerves are embedded in the tissue. Samkoe compares it to putting a flashlight under your chin and seeing your cheeks glow. "The reason they glow red isn't because there is blood inside them," she explains. Rather, it has to do with the relative wavelengths that make up white light; since red wavelengths are the longest, they go through tissue more easily, lighting up your cheeks with spooky red light.  

By the same principle, she can excite fluorophores with multiple wavelengths of light and then look at the mathematical relationship between them to determine how deep they are within the tissue. Those data are then translated into a "color map" on the screen showing relative depths of objects that a surgeon can use to guide robotic implements.  

The end goal is a 90-percent reduction in nerve injury during surgery—results that could dramatically improve a patient's quality of life. "We're focused on prostate cancer, but this is widely applicable to any surgery," Samkoe adds. The same sorts of damage are observed in gynecological and gastroenterological surgeries, for example. "The wonderful thing about the ARPA-H project is that it's bringing together large teams of people to attack a problem from many different directions," Samkoe says.  

Samkoe is primarily focused on the chemical basis for the mapping, while Halter and fellow engineering professors Keith Paulsen and Scott Davis tackle electronic imaging. "Kim's very thoughtful and interested in the work we are doing," says Halter, who appreciates how collaborative Samkoe has been in adapting her techniques to produce better imaging. "She listens and assimilates opinions from other colleagues to develop more impactful technology beyond just her focus." 

In addition to Thayer and DHMC, the team is collaborating with researchers from Johns Hopkins University and Oregon Health and Science University as well as companies including Intuitive Surgical, Trace Biosciences, and Dartmouth Engineering startup QUEL Imaging. "We're bringing together leaders in optics, instrument design, image technology, and molecular agent design," says Samkoe. "These technologies can often take a long time to produce—hopefully funding like this will help bring solutions to the clinic more quickly."  

Photo by Rob Strong '04

Personalized Therapies 

Along with applying fluorescence to surgery, Samkoe has been working on techniques to improve cancer treatments involving molecular therapies. Those therapies work by injecting agents into the body to target proteins or other molecules within cancer cells, disrupting their growth or function. "Cancer cells are normal cells gone awry," Samkoe says, "so they are not doing what they are supposed to anymore."  

For example, cancer cells might hijack normal cell processes to sprout more growth receptors on the outside of a cell, leading to uncontrolled cell growth that causes a tumor. "A normal cell might have 1,000 but a cancer cell has 1,000,000," Samkoe explains. A molecular therapy might target those growth receptors, blocking them to turn them off like a switch, stopping cancerous growth. The challenge? These proteins are hard to target since they are naturally occurring molecules within the body and may vary greatly from patient to patient or even cell to cell. 

"Oftentimes to know whether a person is eligible for a therapy, you have to take a biopsy of the tumor to test if a person has a protein a drug is targeted to," Samkoe says. Her methods can help diagnose cancer cells without surgical intervention by attaching fluorophores to molecules that bind to these proteins, lighting them up like bowling pins in a cosmic alley to show where to strike. Once identified, those proteins can be targeted where they occur.  

Samkoe has even created ways to better determine proper dosage, in part by adapting techniques she first developed as a postdoc with fellow postdoc Ken Tichauer, now associate professor of biomedical engineering at the Illinois Institute of Technology. "Inherently, fluorescence is a qualitative measure that can tell you whether something is there, but not really how much there is," Samkoe says. She and Tichauer have cleverly worked around this limitation by using fluorophores with two different colors simultaneously—one that binds to the protein and one that doesn't.  

By injecting both at the same time and shining a light on them, researchers can monitor how the colors change, showing how much of the fluorescence is being bound and how much is being washed away. "Dividing one by another shows us how much of the target is present," Samkoe says. "These techniques can show us how much is there, how long it is there, and how strongly it is binding—all information relevant to therapy."  

With her chemistry background, Samkoe has spearheaded the molecular basis for the techniques, while physicist Tichauer focused on the mathematical modeling. Together, they've collaborated on nearly 50 papers during the past 13 years. "Kim is one of my favorite people," he says. "She's brilliant but also generous and extremely collaborative—she makes the hardest of science fun to work on." Through the years, he's learned to stay quiet in the face of obstacles and watch Samkoe work out a solution. "She's an outstanding problem solver. She's always looking for the right things to do—not just blindly solving issues but thinking about how to get technology into the clinic."  

One of the challenges is that molecular therapies are so individual to patients that these potential solutions often fail in clinical trials due to difficulty in determining a common dosage. "Many new and exciting cancer therapies fail in clinical trials because of this variability and our inability to repeatedly test the cancer before and during treatment due to the need to remove tissue samples or undergo imaging with potentially harmful side effects," Samkoe explains. 

Using similar techniques with two differently colored fluorophores, she has experimented with safe and noninvasive ways to show in real time how effectively a therapy is working, monitoring patients across days or even weeks to see if the amount of a particular target decreases. "Sometimes medicine doesn't get to the cells it’s supposed to because of the particular characteristics of a tumor. Because we have this ratio information, we realized we can actually use it to visualize therapeutic agents binding in real time," says Samkoe, who recently published the findings in the journal Molecular Pharmaceutics with Tichauer and other colleagues, including Thayer's Xiaochun Xu and Sassan Hodge and Geisel's Yichen Feng. 

By monitoring real-time changes in the tumor, she says, a doctor could gauge effectiveness much more rapidly rather than waiting weeks or months to see if a tumor changes in size. "Every tumor is different, and every patient is different—sometimes even every cell in the tumor is different," she says. "Instead of giving every patient the same dose of medicine, the idea is to figure out what medicine is going to work best for that patient and how much to give them."  

Using these fluorescent imaging techniques could lead to the holy grail in medicine—individualized treatments that customize therapies and dosages to target the proteins expressed in a particular patient or tumor. By more precisely attacking cancerous cells, clinicians increase the likelihood they can defeat the cancer for good. "From a scientific standpoint, this is exciting because we are developing new methods to visualize and assess biological processes that could have far-reaching effects," Samkoe says, "improving not only clinicians' ability to treat patients, but also patients' overall health and longevity."