Skip to Content

A Sharper Picture of Health

Powerful new imaging technologies pinpoint the molecular events involved in diseases, promising a safer alternative to biopsies.

Northwestern University chemist Thomas J. Meade means no disrespect to his medical colleagues, but when he looks at the state of the art in diagnostics, he suggests that, for some procedures, physicians might as well use “stone knives.” Take, for example, mammography. “You know going in that there’s a one in five chance of a false positive or a false negative. You have an x-ray that’s not even smart enough to differentiate a shadow cast by a calcium spot from a tumor. After reading the film and seeing a shadow, they do the prudent thing and stick a 16-gauge needle in you for a biopsy. Then you have to spend the next five days freaking out that you’ve got breast cancer until you get the results,” he says.

Some of Meade’s annoyance stems from the fact that his wife endured just such a false alarm-as will an estimated quarter to half of all women who undergo annual mammogram screening over the course of 10 years. But Meade’s criticisms go well beyond the specific failings of mammography and breast biopsies; to him and a growing number of other medical researchers, today’s diagnostic tools are too uncertain and invasive-just too primitive. Working at select academic centers and industrial labs around the world, these researchers are developing a suite of new tools that will enable doctors to spot disease instantly and accurately, without ever taking a scalpel or biopsy needle to their patients’ skin.

The new discipline is called “molecular imaging,” and it is fundamentally altering physicians’ ability to view the body and its processes. Most conventional imaging tools, from x-ray to magnetic-resonance imaging, provide anatomical or structural information: is there a lump in the breast or a shadow in the lung? Molecular imaging goes beyond anatomical information to reveal functional data-the cellular activities that characterize tumor growth or inflammation, for example.

This is important because cancer and other diseases often begin with subtle cellular changes, well before a structural abnormality, such as a tumor, is detectable. What’s more, the new advanced imaging methods can help distinguish between diseases that look similar but actually involve different molecular malfunctions-and thus require different treatments. “Disease is being redefined in terms of its molecular signature,” explains Daniel Sullivan, associate director of the National Cancer Institute and head of the institute’s biomedical-imaging program. “In the future, people will talk about cancer by the molecular abnormality, not by the organ of origin.”

And in the future, molecular imaging could be integral to every step of health care. Exquisitely sensitive periodic scans could flag any worrisome changes-the presence of a particular protein associated with the beginning of a cancer, for instance. Should doctors eventually spot a lesion, the imaging process itself would yield enough information about the biochemical malfunction not only to make the diagnosis without a biopsy, but also to help determine the best therapeutic option. Still more noninvasive imaging would closely track the treatment’s progress and the course of the disease.

It might take 20 years or more for this complete picture to emerge. But the first generation of technologies to make it possible is already appearing. Thanks to advances in a range of disciplines, from molecular biology to optics to computation, researchers have begun to design chemicals that, once injected into the body, swarm to particular molecules associated with certain diseases and light them up, allowing physicians to easily spot problem areas.

More than half a dozen such molecular-imaging agents are now on the market, most of them for cancer diagnosis; another handful are in clinical trials, and even more are in the development pipeline. Targets for these new diagnostic tools include not only cancers but cardiovascular disease and ills of the central nervous system. “All these technologies are growing incredibly rapidly,” says Harvard Medical School radiologist Umar Mahmood, a principal investigator at Massachusetts General Hospital’s Center for Molecular Imaging and Research. “We’re not doing incremental work. These are leapfrog advances.”

Indeed, it’s a big enough jump forward that government, the established diagnostics industry, and a few venture capitalists are making substantial investments in the field. The National Cancer Institute has designated molecular imaging an “extraordinary opportunity,” spending more than $100 million on it in recent years and asking for $78 million in 2004. To keep up, companies like General Electric-which already has a $9 billion business in conventional medical imaging machines and systems-are venturing further into molecular biology and chemistry and inking deals with major pharmaceutical companies and startups to develop new imaging agents (see “A Snapshot of Molecular-Imaging Companies,” bottom). For the industry and patients alike, says Eric Stahre, general manager for genomics and molecular imaging at GE Medical, “molecular imaging has the potential to change the game.”

A Snapshot of Molecular-Imaging Companies
COMPANY
PROJECT PARTNERS
GE Medical Systems
(Waukesha, WI)
Imaging instruments and agents GlaxoSmithKline, Amersham Health
Kereos
(St. Louis, MO)
Nanoparticle imaging agents for cancer and heart disease Dow Chemical, Philips Medical Systems
MetaProbe
(San Diego, CA)
Imaging probes for neuro- degenerative, cardiovascular,
and liver diseases
University of Wisconsin, Vanderbilt
Molecular Insight
Pharmaceuticals
(Cambridge, MA)
Imaging agents designed to identify heart damage Harvard University, Syracuse University, Georgetown University
Theseus Imaging
(Worcester, MA)
Imaging agents for cancer and heart disease Philips Medical Systems
Xenogen (Alameda, CA) Optical imaging in animal models of disease AstraZeneca, Biogen, Bristol-Myers Squibb, Chiron, Eli Lilly, Millennium, Novartis, Pfizer

The Next Blockbusters

Take Apomate, a molecular-imaging agent made by Boston, MA-based Theseus Imaging (a subsidiary of North American Scientific) that gives physicians a novel view of a biological drama that plays out in the body all the time. The complex chain of events that leads to a cell’s death, a process that biologists call apoptosis, is central to everything from embryonic development to aging; Apomate allows researchers to directly observe these events. By imaging cells’ death throes, doctors might be able to see if a particular chemotherapy is successfully killing tumor cells, say, or more accurately assess damage caused by a heart attack.

Like many of the other new imaging agents in development, Apomate draws on the recent explosion in biologists’ understanding of the details of the body’s molecular processes. Among other findings, scientists have begun to unravel the precise details of apoptosis. It turns out that dying cells expose a binding site that’s normally concealed. A naturally occurring protein then binds to that site, marking the cells for destruction by the immune system. Theseus created Apomate by engineering a synthetic version of the protein and linking it to a radioactive isotope that shows up under a scanner. When a patient is injected with Apomate, areas of the body where many cells are dying light up.

Apomate’s potential market is large because of the myriad roles apoptosis plays in the body and the many ways doctors could use the agent. In human trials-ongoing in Europe now and likely to begin shortly in the United States-doctors are giving Apomate to lung cancer patients shortly after their first injections of chemotherapy. The aim is to determine within a day or two of the injections whether the drug is actually killing tumor cells. Patients who aren’t helped by a particular chemotherapy drug can then be spared the wasted time and often debilitating side effects of the treatment-and can more quickly move on to explore other options. “Only 20 percent of these patients respond to the therapies, but 90 percent have side effects,” explains Allan M. Green, chief technology officer at North American Scientific and Theseus. “To recognize early responses patient by patient is the most important near-term contribution we can make.”

Theseus has also tested Apomate in more than 50 heart attack patients. Green says investigators have discovered a small subset of patients in whom apoptosis continues to take place in heart tissue even months after an attack. And researchers suspect these patients may be the most likely candidates for subsequent heart failure. If Apomate could help identify these patients, whose heart cells are continuing to die off, it could provide valuable clues as to who should be treated most aggressively-before their hearts begin to fail.

Researchers have also been using Apomate to try and identify unstable plaques in coronary blood vessels. It appears that the plaques most likely to rupture and cause heart attacks demonstrate a measurable amount of apoptosis as they crack and chip. If this work on imaging vulnerable plaques pays off, it could help doctors identify ahead of time some of the hundreds of thousands of people each year whose first indications of heart disease might otherwise be lethal heart attacks.

Doctors currently have ways of looking at structural changes in coronary arteries, but to find earlier danger signs “you need to know the biology,” says Stanford University radiologist Francis G. Blankenberg, a consultant to Theseus. Blankenberg says he envisions imaging schemes that look for apoptosis becoming part of a standard battery of tests for patients showing any kind of chest pain.

Noise Reduction

Despite its potential, however, molecular imaging is not without its technical challenges. For one thing, researchers are working hard to ensure that the images it produces are clear. Imaging agents are always on-always emitting radioactivity, like Apomate, or always glowing, in the case of some others. That works fine when there is an abundance of target molecules for the agents to bind to. The trouble is, sometimes the molecules characterizing a problem are so scarce that the few imaging agents that do reach and bind to them are lost in a haze generated by unbound agents floating nearby. That makes it hard to pick out a few precancerous cells in an otherwise healthy organ, for instance. And other times the agents collect in locations such as the liver, where they give off a bright but meaningless glow.

To address these problems, Northwestern’s Meade and other researchers are inventing chemistry designed to keep the imaging agents invisible until they find their target molecules. “We’re making molecular beacons that respond to physiological conditions,” explains Meade. “They’re off when injected, and only turned on by the presence of an enzyme target.” That simple-sounding goal will nonetheless demand considerable basic research and complex chemistry.

Meade, for one, has come up with a novel scheme in which a target enzyme chews off the equivalent of a cap on the imaging agent, allowing it to beam out its signal. He hopes the approach will allow diagnosticians to use high-resolution MRI and computed-tomography scanners to probe ailments such as stroke, schizophrenia, and Alzheimer’s disease.

Researchers at Massachusetts General Hospital are putting some of these same principles to work to improve the diagnosis of colon cancer-the second most common cause of cancer death in the United States, with more than 50,000 fatalities each year. Colonoscopy has helped physicians find colon cancer earlier. However, it can have difficulty differentiating between dangerous and more benign polyps. In a lab at Mass. General headed by Ralph Weissleder, scientists have found that an enzyme called cathepsin-B appears in higher concentrations in the most dangerous polyps than it does in nearby tissue and in other polyps.

Armed with this biological insight, the researchers designed a clever new imaging agent, taking advantage of the fact that cathepsin-B is an enzyme that cuts up specific proteins. They constructed the agent out of a fluorescent protein fragment attached to another molecule that keeps the fluorescence quenched. When the imaging agent finds cathepsin-B, the enzyme cleaves off the quenching arm, freeing up the probe to glow brightly. Since targets like cathepsin-B exist in very small quantities, turning off extraneous unbound agent molecules is like darkening the stars in a night sky to better spot a passing comet.

Because light scatters as it penetrates deep into tissues, it would be difficult to scan a human patient’s colon optically from outside the body. But the technology could be paired with conventional colonoscopy. In the future, doctors may inject a patient with a fluorescent imaging probe designed to find the enzyme, and then examine the colon using a fiber-optic scope that picks up the fluorescence (see “Molecular Colonoscopy,” below). This will allow doctors to distinguish between the different kinds of polyps in real time, with a minimally invasive approach, instead of having to cut out sample tissue and send it to a pathology lab, then wait days or weeks for the results.

Molecular Colonoscopy

To distinguish between dangerous and benign polyps, doctors might inject a molecular-imaging agent targeting an enzyme that’s concentrated in dangerous polyps. The enzyme itself would activate the agent by cutting it up and releasing particles that glow when light from a fiber-optic scope shines on the colon wall. (Illustration by John MacNeill; source: Umar Mahmood)

As a next step, says Mass. General’s Mahmood, the researchers are now performing test colonoscopies on lab mice using the new imaging agent. According to Mahmood, the technique could make it possible to avoid many colon biopsies in five to ten years.

In the near term, optical imaging could also help improve the accuracy of breast biopsies. These procedures now can miss malignant cells because it’s difficult for physicians working off of two-dimensional mammograms to know exactly where in the breast to place their needles. Experts estimate that in the United States alone, some 50,000 to 100,000 breast biopsies each year don’t find existing cancer cells-and so do not properly diagnose the women’s cancers.

To address the problem, University of Wisconsin-Madison biomedical engineer Nimmi Ramanujam is exploiting the fact that biological tissues naturally fluoresce in response to stimulation by certain wavelengths of light-and that healthy and cancerous tissues fluoresce differently. Ramanujam scaled down optical imaging technology to create a tiny fiber-optic sensor that can be threaded right through a biopsy needle. When the doctor inserts the needle into the breast, the device sends light into the tissue and collects the fluorescence emitted by cells at the needle’s tip; algorithms developed by Ramanujam analyze this fluorescence in real time to distinguish between the telltale optical signatures of healthy and cancerous tissues.

Ramanujam and her colleagues at the University of Wisconsin Medical School are already testing the technology on women undergoing breast cancer surgery and plan trials with women undergoing breast biopsy within the next year.

Blurred Boundaries

These rapid advances in molecular imaging are helping to blur medicine’s traditional boundary between diagnosis and treatment. That’s because the potential to pinpoint molecular events involved in a disease, which is at the core of new imaging methods, raises an even more tantalizing possibility: in addition to diagnosing the problem, why not actively disrupt the process while you’re at it?

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.