ONE of the prospects supporters of the Human Genome Project held out was personalised medicine. Knowing which genes were involved in a particular patient’s disease would allow drugs to be deployed with greater precision.
That is starting to happen in the field of cancer. Several targeted therapies, aimed at specific cancer-causing mutations, including Gleevec for chronic myelogenous leukaemia and Herceptin for some types of breast cancer, have been spectacularly successful. Yet in most cases of cancer doctors still base their treatment on where in the body a tumour has sprung up, rather than on which molecular aberrations have caused it.
The same is true of medical researchers recruiting volunteers for clinical trials, especially those known as phase I trials, in which a new drug is tested on people for the first time. Participants in such trials are often those whose tumour has spread beyond its original site, and will probably prove fatal. Usually, they have tried all proven therapies, to no avail. Their precarious condition means they are rarely accepted for phase II and III trials, which are more complicated and extensive.
Oddly, though, even if the drug being tested is a targeted therapy the tradition in phase I trials has been to gather together patients with, say, lung cancer and assume that all carry the relevant mutation. That is because such trials are concerned mainly with testing a drug’s safety, not its efficacy. The volunteers are usually happy to go along with this. But the odds are not good. On average, fewer than 5% of participants in phase I trials respond successfully to the treatment.
Many scientists think this approach is a wasted opportunity. They believe that matching volunteers’ genetic profiles to the drugs being tested will not only be better for the volunteers, but may also speed up the trials, and save millions of dollars in the process. One such is Apostolia-Maria Tsimberidou of the University of Texas’s MD Anderson Cancer Centre, in Houston. And her preliminary results, presented this week at a meeting of the American Society of Clinical Oncology in Chicago, suggest she is right.
Dr Tsimberidou and her colleagues selected volunteers with late-stage cancer whose tumours were caused by single, known mutations and did the equivalent of a phase I trial on them. In the case of the 175 volunteers for whose mutations a targeted therapy existed, it was given. The remaining 116 received traditional treatment. Of those in the targeted-trial group, 29% responded positively to the therapy. (The tumours of four disappeared altogether.) Only 5% of those on untargeted drugs showed improvements—precisely in line with the historical figure. Those on targeted drugs also responded for longer than those who were not (five months, as against two months, on average). And they lived for 13 months compared with nine.
Dr Tsimberidou cautions that the data, though encouraging, need to be qualified. Since her patients were not assigned to the targeted and untargeted groups at random, the gold standard for clinical trials, some hidden variable, rather than the treatment, may explain the different outcomes. The untargeted group might, for example, have been sicker to start with. They did, indeed, have higher initial levels of lactate dehydrogenase, an enzyme associated with tissue breakdown, so this possibility should not be discounted. However, those on targeted therapy did better, compared with their reactions to previous treatments, than those who were not.
Not surprisingly, drug companies are taking an interest in the findings. If a hefty proportion of patients in an early trial show a robust response to a drug, its developer may be able to skip phase II, the first real look at a drug’s pharmacological credentials, and jump straight to a phase III trial—the ultimate test of efficacy. This might save a company several years of clinical studies, and an awful lot of money.
Widespread adoption of the new approach may, however, be difficult. Tests to detect cancer-causing mutations are still in their infancy. Many such mutations have not yet been identified, and even when a known mutation is found in a tumour, drugs to block its effects may be unavailable. On top of that, many patients will not participate in a trial that requires the collection of an additional sample from their tumour, for this is an invasive, painful and potentially dangerous procedure. Indeed, one in ten of the applicants for Dr Tsimberidou’s trial was unable to provide samples which could be tested for their molecular make-up.
However, Johann de Bono and his colleagues at the Royal Marsden Hospital, in London, think that they have found a solution to this problem. Rather than probe tumours directly, they can test bits of tumour DNA that float in the bloodstream, and identify genetic mutations in this way with nearly the same accuracy as the direct method provides. If these preliminary results are confirmed, more patients with advanced cancer will have been offered a tantalising glimmer of hope.