Editor’s Note: This story was written by Bryan Kohrs, a junior in Biophysics at Johns Hopkins University with a strong interest in science writing and science. It first appeared in the 2013 issue of Nano-Bio Magazine.
Denis Wirtz lines up next to many other scientists in the war on cancer. But while others battle with familiar technologies and ideas, Wirtz has armed himself with a new imaging technology, a fresh strategy on how to better combat this dreaded disease and a five year grant from the National Cancer Institute (NCI).
Wirtz, a professor of Chemical and Biomolecular Engineering at Johns Hopkins, is carving out a niche for himself in cancer research by focusing on the look and physical structure of cancerous cells rather than the genes from which the cells originated. He believes that this technology will not only help doctors predict how cancer progresses, but will eventually change the way cancer is treated from a therapeutic standpoint. The technology is the keystone of the new Johns Hopkins Center for Digital Pathology, which Wirtz directs.
To understand how this technology works, consider this analogy: A cell is like a Lego brick. Just as individual Legos can come together to create a building, millions of different cells come together to make a human being. Certain bricks serve different purposes in a building in the same way that distinctive cells carry out certain functions in the body. The properties that make some Legos better suited for one purpose over another are their size and shape. For example, if a piece is flat and wide, it should go on the base. Structural characteristics that make cells unique include overall size, shape, the size and shape of the different cell parts or organelles, the composition of certain organelles, and hundreds of other parameters.
Wirtz’s technology uses a modified scanning electron microscope and a process called high-throughput cell phenotyping (HTCP) to instantly make hundreds of thousands of highly specific measurements defining each of these structural cellular characteristics of each single cell on a slide. Wirtz has software that uses an algorithm that adds up all of the different measurements and gives a cell a structural “score,” which quantifies the look of the cell with a number. The process will be automated and will take just minutes for a slide of cells to be analyzed and given an overall structural score, which averages the scores of all the cells on the slide.
Anirban Maitra, a professor of Oncology and Pathology at Johns Hopkins School of Medicine who is collaborating with Wirtz on this project, explains the benefits of automating this process, “If you were looking at a cell with the naked eye, you would say it has a large nucleus, medium sized nucleus, or a small nucleus. What automation allows you to do is to spread that crude three-tiered category into hundreds of small denominational events that you could then objectively add up and get a score.”
Over the course of the next five years, Wirtz plans to use HTCPanalysis as a clinically applicable tool that can help doctors treat cancer patients with more personalized therapies.
“Currently, we have a very crude approach to therapy even with the targeted therapies that are being developed. The vast majority of patients in cancer care and oncology get what are called cytotoxic agents, the old agents that were made many years ago,” said Maitra. But by using HTCP to see how cancers that look a certain way respond to certain treatments, doctors will be able to better personalize cancer treatments.
To make the project clinically applicable, Wirtz, with the help of Maitra and Ralph Hruban, also a professor of Oncology and Pathology at Johns Hopkins School of Medicine who is collaborating with the team, will be working to create the first “phenotypic database,” or a cell-feature-focused database. It will combine patient data like age, sex, cancer type, progression, treatment used, genetic sequencing results (analysis of tumor from a genetic standpoint), and so forth in an online, “cloud” database and then also add in the structure score of the patient’s tumor performed from HTCP.
At the moment, Hopkins is the only university with Wirtz’s new technology. The plan is for hospitals across the nation to begin uploading patient information to the database online and sending slides of cancer tumor cells to Hopkins or an alternate research facility using this technology. There, independent researchers will analyze the cells and add the HTCP analysis to the patient information online.
Doctors can upload all of this data into the cloud and help the database grow initially. Eventually, an oncologist in Chicago, treating a 70-year-old man with lung cancer, and a HTCP score of X will be able to go online and find that there were two similar patients, a 65-year-old man with lung cancer in Baltimore and a 75-year-old woman with lung cancer in California, both with a score of X as well. The physician would discover that the man in Baltimore was treated with chemotherapy A and died in six months, while the woman in California was treated with chemotherapy B and was cured. Doctors will be able to make more informed treatment decisions.
Classifying the morphological characteristics of cancer is a shift from the traditional genetic approach to categorizing cancer cells. Previously, scientists researched cancer from a genetic standpoint, linking specific genetic mutations to specific cases of cancer. While this has lead to gene-targeted therapies, Wirtz wants to take a different approach to cancer research. He wants to look beyond the genetic origin of cancer and focus on what cancerous cells look like.
“We’ve come to realize that it is the heterogeneity – the diversity of cells that have different characteristics – is also important in evaluating a cancer case. In the end what matters are the cell properties.
That’s what we measure,” Wirtz explained. The rationale for this new approach, Wirtz explained, is that while cells can be identical genetically, they can vary tremendously in structure, just as two identical twins can develop to be very different people, both physically and personally. Cells from one tumor could become metastatic, latch onto a new organ, and start a new tumor that eventually kills the patient, while a genetically identical set of cells could remain localized and die as soon as they detach from the original tumor.
Wirtz’s theory is that the key to cancer treatment prognostication lies not in cancer genetics, but in the physical attributes of cancerous cells. For example, you could say that muscle definition and physical fitness would be strongly correlated with athletics and would therefore be able to be used to predict who out of twenty people would become athletes. Wirtz believes that his technology will allow doctors to do the same thing with cancer.
With this new technology Wirtz hopes to figure out what triggers cancer cells to metastasize. For example, do small, elliptical cells with large nuclei metastasize better than large, rod-like cells with small nuclei? He explained that cells that metastasize have to be super-cells, much like super-heroes are better, faster, and stronger than other humans.
“Millions of cells are shed by tumors every day, but only one or two of them will have what it takes to become metastatic. These are the decathlon cells. We need to figure out what the physical properties are that give these cells an edge,” Wirtz says.
Maitra poses the question that guides the project in its applications towards therapeutic cancer treatment, “We have a lot of different drugs out there right now. Some work, some don’t. The problem is you only find out if they worked retroactively. You give it to a patient and six months later the metastasis keeps growing and you know if it’s worked or not. But wouldn’t it be nice if we knew going into the treatment that these patients would respond to a particular regimen and these other patients respond to another regimen?”
Maitra believes that conceptually, this project is paradigm shifting. “Wirtz is analyzing cancer in a brand new way. Extending this tool into an open-access cancer database, the project seems to have a bright future for helping doctors treat patients.”
Maitra makes sure to keep the project in perspective while being hopeful about the direction of this project, “It is very preliminary at this point. We have a long way to go before we can actually say this is a clinically applicable technology, but what we are doing right now is working our way up there.”