Current neuro-rehabilitation treatment relies on experience-dependent plasticity, one of the most basic concepts in neuroscience. Our brains are surprisingly plastic, meaning they have the capacity to rewire and form new neuronal connections throughout life, and experience-dependent plasticity refers to the process by which our brains do this based on our experiences. This phenomenon was first identified by a scientist named William James in the late 19th century, when he observed changes in neural pathways linked to the formation of habitual behaviors.

Neuroscientist Mike Merzenich discovered a striking example of such plasticity in the 1980s, when his lab found that removing the finger of an adult monkey induced large-scale remapping of the neurons in its motor cortex. The motor cortex is a region of the brain involved in motor processing, helping plan and move different parts of the body. Merzenich's team found that the neurons in the monkey's motor cortex that were responsible for moving the lost finger rewired themselves once the finger was gone — establishing new neuronal connections to help adjust to its loss and to move the remaining fingers.

We know that after strokes in humans, the remaining healthy neurons in our brain have the same ability to rewire, strengthening intact neural connections. This incredible ability means that brains can be retrained after strokes to potentially restore impaired functions. And a patient's behaviors and experiences play a key role in the process. Existing neuro-rehabilitation approaches try to capitalize on this ability, for example, to help stroke survivors re-learn how to brush their hair.

There’s just one catch: The type of behavioral experience matters. Not all experiences are ‘good’ for brain plasticity. Unfortunately, the brain can also rewire itself in a counterproductive manner if patients provided the wrong kind of experience after a stroke. For example, a previous study found that training rats to compensate for loss of function on one side of the body by using their other, healthy side for a given task after a stroke actually worsened their recovery on the injured side — even more so than no rehabilitation at all. Training the uninjured side actually rewired the brain in a manner that supported movement with this side, and hampered the recovery of the injured side.

What does this mean for clinical stroke patients? It turns out learning to compensate with the non-injured side actually decreases the possibility that survivors will regain function in their injured side. Therefore, neuro-rehabilitation strategies that don't limit or restrict use of the healthy side can actually hinder long-term recovery.

In humans, the answer as to how to implement these scientific findings is more complicated than in rats, for which researchers simply encouraged use of the injured limb. Of course, from a clinical research standpoint, it is very challenging to force stroke patients to only use their injured side. Many patients understandably feel that the easiest way to resume some type of normalcy after stroke is to learn how to accomplish daily activities by compensating with their non-impaired side. Not being able to perform tasks like getting dressed, eating, and bathing that they once took for granted is extremely frustrating.

However, most patients are unaware that learning to compensate can have a detrimental impact on long-term recovery of the impaired side. This may seem less important to stroke survivors early on, as compensation allows them to dramatically and rapidly improve their quality of life. But it becomes more important later, when this compensation diminishes any chance of recovery of the injured side.

Early efforts to combat the negative effects of compensation can involve extreme measures, including therapies like Constraint-induced Movement Therapy (CIMT), which restricts usage of the uninjured side in patients for at least six hours per day (and in some cases up to over 90 percent of the waking day). Although CIMT has had success in clinical trials, it can feel impractical for stroke survivors. In fact, the major issue with early CIMT therapies was patient compliance.

In reality, some compensation is both needed and expected for stroke patients with upper-extremity impairments to recover. Compensation isn’t all bad — especially if it helps patients regain some type of normalcy after stroke. A recent meta-analysis found that modified versions of traditional CIMT, which involved longer durations of less-intensive therapy, were more successful than traditional rehabilitation strategies.  

Finding the right combination of compensation and task-specific training in stroke survivors is a work-in-progress. But there's an increasing awareness of just how powerful post-stroke experience is in reshaping the damaged brain — and that is helping patients move toward better outcomes. 

Thousands of miles away, researchers at the University of Pittsburgh were racing to refine a therapy that would ultimately save the teenager's life. The results of Carnell-Holdaway's treatment were published recently in the journal Nature Medicine; now, the treatment is gaining renewed interest, with some even going as far as to call its effects "miraculous." 

The treatment, called phage therapy, has existed for a century, and its underlying concept is elegant: Co-opt viruses that naturally infect bacteria to treat infections while ignoring human host cells. The use of these viruses, known as bacteriophages or phages for short, proliferated in the Soviet Union in the early 20th century in the absence of antibiotics, but the practice has only recently been rediscovered. It is gaining popularity in the United States and Europe, where it is being tested as a way to treat everything from burns to urinary tract infections. 

As a last-ditch option, one of Carnell-Holdaway's doctors reached out to Graham Hatfull, a professor at the University of Pittsburgh whose lab studies bacteriophages that infect the tuberculosis family of bacteria. Carnell-Holdaway was infected with a member of this bacteria family, which is known to target immunocompromised patients, and Hatfull and his team searched through a library of over 10,000 different phages to find the ones that would attack her strain the best. After identifying three that could infect the bacterium effectively, the researchers had one more roadblock to surmount: One of the phages was programmed to infect, but not kill the bacterium. This behavior is known as the lysogenic cycle, in contrast to the lytic cycle in which a virus copies itself before bursting out of its host cell and destroying it.

To address the issue, Hatfull's team genetically engineered the misbehaving phage to express lytic behavior. This was the first time that an engineered phage has been used in a clinical setting. After conducting the necessary quality control experiments and obtaining regulatory permission to ship the phages from the U.S. to the U.K., the team sent off the three-phage cocktail to Carnell-Holdaway's doctors, about six months after they had been first contacted.

But while time pressures were ever-present in the back of the researchers' minds, Hatfull told me in an interview that the work began as a purely hypothetical exercise — seeing if their library contained the right phages — and only later morphed into a treatment Carnell-Holdaway might actually receive.

Carnell-Holdaway started to show improvement within weeks of beginning intravenous treatment with the phages every 12 hours — skin lesions caused by the infection started to clear up, her lung function improved and she gained weight. Still, since her treatment was not conducted like an experiment, it is impossible to be 100% certain that the phage therapy caused her recovery.

After nine days of monitoring in the hospital, Carnell-Holdaway had not reported any adverse reactions to the phages, an outcome "that obviously came as a not insubstantial relief to all of us," Hatfull says. When her doctors decided to discharge her from the hospital again, they were optimistic about her chances of recovery.

Even with the long history of phage therapy in Eastern Europe, this was the first recorded instance of both a genetically engineered phage used clinically and an infection in the tuberculosis family being treated with phages. 

"We're in uncharted territory," Hatfull says.

But that has not stopped researchers and media outlets alike from speculating that phage therapy may one day replace antibiotics as first-line treatments of drug-resistant infections. This kind of "silver bullet" thinking is misplaced, says Hatfull, though he does think phage therapy will likely be the solution to some types of highly resistant infections.

The story of Carnell-Holdaway's treatment is a case study into the difficulties facing widespread clinical implementation of phage therapy and a potential roadmap to solving them.

One problem is the phage's specificity. From an evolutionary standpoint, it makes sense that a highly specialized phage would perform better than a jack of all trades. This has resulted in highly specific phages that are best at infecting substrains within strains: The three phages Hatfull's team identified as candidates to treat Carnell-Holdaway's infection were not effective against other isolates of the same bacterium. At the same time, only a tiny fraction of the estimated ten nonillion, or 10³¹, phages have been discovered and categorized in a library. And once these massive libraries are established, there needs to be infrastructure in place for quickly and cheaply determining the few relevant phages for any specific infection.

Programs like SEA-PHAGES, run by Hatfull's lab and the Howard Hughes Medical Institute in the U.S., are trying to address the latter challenge using crowdsourcing. SEA-PHAGES trains undergraduate researchers from more than 100 colleges and universities around the country to discover and sequence phages over the course of two semesters, and the program will soon reach 3,000 sequenced phage genomes.

Carnell-Holdaway will soon reach a milestone, too: One year of continuous phage treatment. During that time, she has returned to a normal life and is busy with schoolwork and A-levels. Hatfull says there is no telling how long Carnell-Holdaway will remain on treatment, nor whether the phages will keep the infection at bay permanently.  

Nevertheless, future work to engineer and refine phage technology offers great promise for the treatment of disease. "It's really exciting," Hatfull says. "We love this idea of being able to creatively think about how you could fine-tune what you might think of otherwise as a relatively crude technology, and the potential is substantial."

Maddie Bender studies

Microbial Disease Epidemiology