PHOTO: Sirinrath Sirivisoot/Brown University

8 November 2007—Today’s artificial knees and hips let people walk, run, and go about their normal daily lives within months. But they are certainly not perfect. For one thing, they are not as durable as real joints—knee and hip implants typically last 10 to 15 years. And in about 2 percent of patients, the area gets infected and the implants are removed soon after they are put in.

There is no quick, effective way to find out whether an implant is infected or working well, says Thomas Webster, associate professor of engineering and orthopedic surgery at Brown University, in Providence, R.I. Bone-density scans and X-rays are not sensitive enough to pick up the first signs of success or failure.

Webster wants to make smart orthopedic implants that monitor their own progress and then transmit the information, by radio, to a handheld device. If infections occur, the implants could not only alert doctors, but could also release a drug. “Rather than somebody going in to get a bone scan or X-ray months later to see if the implant is working, we want to make an implant that responds intelligently,” he says.

Strong, lightweight titanium is the material of choice for implants. Webster is using a different material, though: titanium with an oxidized surface that’s coated with carbon nanotubes. The nanotubes would essentially work as electrodes. By applying a voltage and measuring the conductivity of the implant surface, doctors could use the nanotubes to sense one of three things that can happen to an implant: healthy bone growth, infection, or the formation of scar tissue. If bone cells are making calcium and depositing bone, they would have a specific conductivity. Similarly, scar tissue and infected tissue would have different conductivities.

By measuring the tiniest change in conductivity, the nanotubes would catch the first signs of an infection. Webster envisions that current pushed through the nanotubes could then degrade a polymer casing on the implant containing antibiotics. Also they could indicate that something is going wrong years after the implant took hold. “If bone is starting to separate, the carbon nanotubes will be able to sense that,” Webster says.

A built-in chip akin to an RFID tag could power the implant using an external RF signal, and it could transmit the implant’s measurements. Doctors would simply hold up a reading device to the patient’s knee or hip to find out if bone is growing or if tissue is starting to get infected.

The researchers’ initial results look encouraging. They have found that carbon nanotube–coated anodized titanium could be a better material for implants than plain titanium. In the September issue of the journal Nanotechnology, they report that up to four times as many bone cells grow on laboratory samples of anodized titanium that is coated with carbon nanotubes than on plain titanium. One reason today’s implants fail is that bone does not grow into and bind very strongly to titanium, so the two eventually separate. The researchers found that bone cells interacting with carbon nanotubes also produce more calcium, which is necessary for healthy bones.

Bone cells might grow better on carbon nanotubes for one of three reasons, Webster says. Anodizing titanium and coating it with carbon nanotubes makes a rough surface with tiny, nanometer-scale features. “These nanostructures emulate the natural roughness of the bones,” Webster says. The titanium used in implants today, in contrast, has bumps and grooves only at the micrometer scale. Another possible reason is that carbon nanotubes, like bone, are conductive, and bone grows better in an electrical current. Also, in the experiments at Brown the carbon nanotubes attracted proteins that are known to improve bone cell growth. Whatever the reason, carbon nanotubes could mean longer-lasting implants in addition to smarter ones.

Among the few others working on smart artificial joints are researchers at Scripps Clinic in La Jolla, Calif. They were the first to implant an electronic artificial knee in a patient, in March 2004. The knee prosthesis contains a wireless strain sensor that measures the compressive forces acting on the implant; a microtransmitter and antenna send the information out to a computer. DePuy Orthopaedics, of Warsaw, Ind., made the implant and MicroStrain, of Williston, Vt., made the sensor. In November last year, Scripps researchers gave a different patient a second-generation version of the knee.

The goal for the Scripps research, however, is not to provide today’s implants with self-monitoring capabilities, according to Darryl D’Lima, director of the orthopedic research laboratories at Scripps. Instead, the researchers want to measure the forces acting on the knee in order to design prostheses with better mechanical properties.

Leading implant manufacturer Zimmer, of Warsaw, Ind., is now working on a smart implant that could sense load, wear, and infection. Zimmer spokesman Brad Bishop says the company’s research is confidential at this point. However, the firm has licensed technology for an orthopedic implant sensor from researchers at Oak Ridge (Tenn.) National Laboratory and the University of Tennessee, Knoxville. The sensor takes a mechanical approach as opposed to the conductivity-based carbon nanotube design. Oak Ridge’s Thomas G. Thundat, who developed the sensor, describes it as a micrometer-size diving board–like structure, called a microcantilever, that measures forces on the implant as well as its temperature. An artificial knee or hip would have three of the microcantilevers to sense load, wear, and infection, respectively. A signal-processing chip would then take the sensors’ readouts and transmit them by radio.

Each of the Oak Ridge sensor’s microcantilevers is about 50 micrometers wide, 200 micrometers long, and 0.5 micrometers thick. It is made of a piezoelectric material, which converts the bending of the cantilever into an electric signal. Force or a change in temperature can displace the end of the cantilevers by a few nanometers or change the frequency at which the cantilever naturally vibrates, so that they produce a measurable signal. To measure infection, Thundat says, you could coat the cantilever with the antibody to a protein or other biomolecule that is formed during infections. When the two molecules attach, the microcantilever’s movement would change.

Smart artificial joints are still years away. Even if Zimmer makes the device, it would have to go through many rounds of tests in patients and government approval. In the end, we might not have artificial knees that are everlasting and perfect, but they might be smart enough to indicate that they are going to fail.