Personalising cancer has led to talk of the 'C' word - cure

Prominent Australian businessman Ron Walker was, in effect, cured by Keytruda, one of the new drugs that have impressed ...
Prominent Australian businessman Ron Walker was, in effect, cured by Keytruda, one of the new drugs that have impressed many doctors. Wayne Taylor

"Cure" is not a word much used by oncologists. The best they normally talk of is "remission".

But the past five years have begun to change that.

More than 70 new drugs have come to market, and describing the consequences of some of them as revolutionary is not hyperbole – at least for those patients lucky enough to respond positively to them.

Being given a diagnosis of advanced melanoma, for example, was once tantamount to being handed a death warrant.

Median life expectancy after such news was six to nine months. But recently developed "immuno­-oncology" drugs, which co­-opt the immune system to fight tumours, are so effective that, in around one-fifth of cases, there is talk among experts that the patients involved have actually been cured.

This sort of upbeat news is reinvigorating the study of cancer. At this year's meeting of the American Society of Clinical Oncology (ASCO), held this week in Chicago, doctors had a spring in their step.

Not only do they have new drugs to deploy, they are also developing better ways of using existing ones.

They are getting better at diagnosis, too, finding methods to study the weak spots of cancers in parts of the body that conventional biopsies cannot reach, and also to pin down tumours that were previously unlocatable.

The upshot is that they are beginning to be able to tailor treatments to the needs of individual patients, an approach called personalised medicine.

These days cancer is seen less as a disease of specific organs, and more as one of molecular mechanisms caused by the mutation of specific genes. The implication of this change of viewpoint is that the best treatment for, say, colorectal cancer may turn out to be something already approved for use against tumours in an entirely different part of the body, such as the breast.

One study presented at ASCO found that 29 of 129 patients responded to drugs that had originally been approved for use on cancers found in parts of the body different from where those patients' own tumours were.

Therapies designed for breast and gastric cancers involving a gene called HER2 were particularly useful. These HER2 drugs act on a growth­-promoting protein that is overproduced in HER2­-positive tumours.

Seven of 20 patients with colorectal cancer, three of eight with bladder cancer and three of six with bile­duct cancer responded well to these drugs.

Metaphysicians

Another study, a "meta­-analysis" of almost 350 early­-stage drug trials which gathered the results of these small experiments together in a statistically meaningful way, tried to work out how much benefit there was in matching the molecular characteristics of the tumour of a patient with his treatment. Such matching proved worthwhile.

Using it caused tumours to shrink by an average of 31 per cent. Established treatments without such matching resulted in an average shrinkage of only 5 per cent.

Work published in the New England Journal of Medicine, to coincide with the ASCO meeting, also showed the value of the molecular approach.

Elli Papaemmanuil of the Memorial Sloan Kettering Cancer Centre in New York and her colleagues have produced a molecular classification of acute myeloid leukaemia. They have divided this disease into 11 classes, each with distinct diagnostic features and clinical outcomes, based on which mutated genes seem to be driving the cancer's development.

While this work has not yet led to better treatments, it seems almost bound to in the future.

ASCO itself sees so much value in the personalised, molecular approach to diagnosis and treatment that, despite it being a professional body for doctors rather than a research organisation in its own right, it has decided to run a clinical trial (its first) to look at this approach's potential.

TAPUR, as the trial is called, will offer patients a genetic test and then select drugs that look to be good matches, but which are not approved for the specific cancer a patient is suffering from. The National Cancer Institute, an American government agency, is trying something similar with a trial it calls MATCH.

This involves sending tumour biopsies to gene­-testing laboratories that then scan them for more than 4000 possible variants of 143 pertinent genes. Indeed, personalised treatment is becoming so fashionable that even America's Vice-President has got involved.

On June 6, Joe Biden announced a project intended to set up a way of sharing genomic and clinical data between cancer researchers, in order to help advance the field.

Taking biopsies such as those that form part of the MATCH trial is a routine part of cancer therapy. It too, though, is ripe for improvement.

Some tissues (blood, lymph and skin, for example) are easy to get at, but many tumours are deep in the body, or in vital organs, or both. Sampling these is invasive and potentially dangerous. Researchers have therefore wondered for a long time whether something as simple as a blood test might replace such a biopsy.

This hope is based on the knowledge that tumours shed pieces of genetic material, known as circulating tumour DNA (ctDNA), into the bloodstream. Until recently, scientific instruments have not been sensitive enough to detect ctDNA routinely and reliably. That is now changing.

"Liquid biopsies", which will not only diagnose hard­-to-get­-at solid tumours but also monitor the progress of their treatment, are on the verge of reality. At the ASCO meeting researchers sponsored by Guardant Health, a diagnostics company, announced the results of one of the largest liquid­ biopsy studies so far.

Liquid asset

This study looked for the ctDNA of six relevant genes in 15,000 patients suffering from one of 50 types of tumour. The test was not perfect. Only 83 per cent of patients had sufficient ctDNA for it to show up.

But in those cases where ctDNA was detected, the mutations indicated were also present in conventional biopsies between 94 per cent and 100 per cent of the time. The test, in other words, is reliable. Moreover, in almost two­-thirds of the cases where ctDNA was detected, the results led to suggestions about how the patients involved should be treated.

If liquid biopsy can be made routine, the clinical consequences will be vast. Conventional biopsies can be both costly and slow to process. Also, the heterogeneity of many tumours, caused by progressive mutation over the course of time, is hard to sample by nipping out one bit of the tumour.

If ctDNA is shed by all parts of a tumour, though, a liquid biopsy will be able to capture these differences. It will, as well, be able to follow them as they progress because, unlike conventional biopsy, it can be done frequently without harming the patient. That is important.

What constitutes the best treatment can change as the tumour itself changes. Many researchers therefore feel it is only a matter of time before liquid biopsies become a standard part of therapy. They are already coming to market.

Foundation Medicine, of Cambridge, Massachusetts, launched a commercial liquid biopsy in May. Qiagen, a German firm, followed suit on June 1. Genomic Health of Redwood City, California, says it will offer a test later this year. And Myriad Genetics, based in Salt Lake City, is also developing such tests.

Such is the excitement over liquid biopsies that some wonder if they might be used to catch cancers even before symptoms are apparent. The earlier a tumour is spotted, the easier it is to cure.

The biggest maker of DNA­-sequencing machines, Illumina, based in San Diego, has said that it will form an offshoot, Grail, to develop just such a test.

The proposed test will use "ultra­-deep sequencing", a technique that reads the DNA in a sample tens of thousands of times over, in order to pick up rare signals such as that from ctDNA.

Early detection

Yet one of the flaws of ctDNA is that it does not reveal where the cancer is in the body. Some argue that MRI is now sophisticated enough to screen individuals for the presence of most cancers.

The Health Nucleus, a firm based in San Diego, is offering full­-body scans using it for just this purpose. David Karow, a clinical radiologist who works both there and at the nearby San Diego campus of the University of California, is optimistic about the potential of the technique for wider use.

He has been part of a study published in Clinical Cancer Research which suggests that a souped­-up form of MRI might become the standard method for prostate­ cancer screening.

His research indicates that such MRI can differentiate between benign and malignant growths, and can distinguish among the latter between those that just need to be monitored, and the "aggressive" ones that need to be treated.

Personalised cancer treatment, long talked of, is thus now becoming real. By detecting problems earlier and getting therapies right first time, it will save lives that might otherwise be lost.

Better knowledge of the underlying processes of cancer, meanwhile, will extend the range of lives that medicine can aspire to save. There is still a long way to go. But gradually and inexorably, the appeals court of oncology is tearing up cancer's death warrants. 

The Economist

Economist