The prospect of developing a virus that selectively kills cancer cells sounds too good to be true.
But a virus that can do just this, at least in the test tube, already exists. The question of whether it will perform the same feat in patients is being tackled at the Institute of Neurological Sciences in Glasgow.
Moira Brown leads one of two groups pursuing this form of cancer therapy.
Her interest began ten years ago when she isolated a variant of herpes simplex, the cold sore virus. It caused no apparent damage to normal tissue but replicated with gusto in the rapidly dividing cells of tumours.
Brown reasoned that if her virus thrived only in malignant tissue, and damaged it in the process, she might have a treatment for cancer.
The initial laboratory work was encouraging. The variant strain - which differs from the regular, or "wild type", virus by a single gene - attacks cells from many types of tumour.
Brown felt confident enough to begin a trial in human patients with glioma, a type of brain tumour.
"The prognosis for glioma is very bad, and nothing really makes any impact on it," she said. The average time from diagnosis to death is a year.
At present, scientists have been engaged in phase-one toxicity trials, not efficacy trials, to find out whether the virus does any unwanted harm. "The answer is that it does not," Brown said.
If the wild-type virus was injected into a patient's brain, the patient would get encephalitis and die.
When Brown's virus is injected into the tumours, there is no sign of such an effect.
The work is categorised as gene therapy and is therefore strictly regulated.
In the first trial of nine patients, the scientists did not even know whether the virus was actually proliferating because they were unable to remove the tumours at the end of the trial.
Three patients are still alive, although Brown has no way of knowing whether this is because of the treatment.
Now the team is running a second trial in which patients are having their tumours removed a week after injection with the virus. The team will then be able to tell whether or not the virus is thriving.
The team is also running a trial in advanced and disseminated melanoma.
Biological control, using parasites, predators and microbes, is a well-established practice in agriculture. The Glasgow work should reveal if it also has a place in medicine.