It is the price of a wonder drug that is the young man's last hope to defeat a tumour that has proved oblivious to radiation, indifferent to chemotherapy and too perilous for surgery - and it costs $8612 a month.

But although the drug is listed on the Pharmaceutical Benefits Scheme, it is not subsidised for the likes of Brad Williams, because it has only been approved for melanoma patients and his tumour is growing in his brain.

"I would have to sell everything I own to be able to afford it," Ms Bantoft said.

"We just don't have that kind of money.

"The frustrating thing is that Brad has just turned 21. It's not like he's a little old man who's had a long life - he's got his whole life ahead of him.

"He is adamant that he doesn't want us to face financial ruin. We just want our son."

Precision medicine - whereby genetic sequencing allows therapies to be tailored to the individual - promises a new dawn in the treatment of rare and fatal cancers.

But the dilemma posed by patients such as Mr Williams is emerging as its greatest challenge, because though the means to treat these people might now be identified, strict regulations surrounding access to PBS medicines renders them unaffordable.

The Pharmaceutical Benefits Advisory Committee approves medicines for government subsidies once the manufacturer can prove they are effective and that they are cost-effective.

There will never be enough patients to run a clinical trial to establish the cost-effectiveness of using a specific drug on a rare cancer, which renders the PBS requirements an insurmountable barrier.

The potential for the drug trametinib to treat Mr Williams arose when his oncologist, Rob Zielinski, sent a sample from the tumour to a German research group for molecular analysis and a mutation was identified that is found in other tumours.

Dr Zielinski said trametinib had produced dramatic results in advanced melanoma patients with this mutation, and drug companies needed to think about cost-share schemes for patients who were not eligible for PBS subsidised medication.

"There's an increasing number of patients who are getting molecular profiling done on their tumours and they throw up these abnormalities that there's an existing drug for but no PBS funding for it," Dr Zielinski said.

"Most companies won't give it for free because they require phase two clinical data, but you're probably never going to get the phase two data because there are so few of these patients.

"The drug may completely not work and I'm not suggesting this it's a life-saving drug, but for a 21-year-old bloke you want to exhaust all possibilities."

Garvan Institute of Medical Research cancer division head David Thomas is attempting to navigate a way through the regulations by bundling rare cancer patients together in a "basket trial" to assess how they respond to targeted therapies.

The approvals process for PBS subsidies was designed around how the drugs worked to control cancer in a specific body organ rather than on a pure molecular basis, Professor Thomas said.

His trial is working to establish evidence that a medicine is effective on a particular mutation, regardless of where in the body it occurred, with each patient prescribed a therapy on the basis of genetic testing.

Many patients were currently spending $5000 to $7000 to have their tumours analysed by direct-to-consumer genetic testing companies based in the United States, Professor Thomas said.

"But they don't get the treatment, they get the test and the problem is, in our country, the range of drugs we can use without paying tens of thousands of dollars - if they are available - is very small," Professor Thomas said.

"In our view it's premature to be offering these tests by direct consumer marketing at high prices.

"The tests themselves cost us less than $1000 to do.

"The purpose of the study is to provide people with not only access to the test in a research context, but also to the treatments which are the real mechanism by which we change people's health outcomes."

Novartis, which manufactures trametinib under the brand name Mekinist, has a compassionate access scheme but its head of government relations, Peter Murphy, said it was becoming more challenging to provide in the current landscape.

Decisions on who could access the scheme were made on an individual basis, based on the evidence and discussions with the treating doctor, he said.

But industry was working with the government to ensure the reimbursement scheme better reflected the way cancer was now diagnosed and treated.

"We want to make sure there's access, but also the government needs to have the right data on these drugs," Mr Murphy said.

"When the data says a drug 'might' work on this particular mutation, that's not the robust data that the government needs."

Oncologist and cancer drug researcher Deme Karikios said the government could consider lowering the threshold for proving a cost benefit in the case of rare cancers, possibly by allowing factors such as tumour shrinkage or symptom improvement to be factored into the analysis rather than relying on survival rates.

Even if genetic sequencing identified a particular mutation in a tumour, there needed to be some evidence an existing medicine could treat it, he said.

"The challenge is whether the government or the PBAC want to use taxpayer funds for those drugs, especially when the benefit is quite modest," Dr Karikios said.

Mr Williams has been given anywhere from six months to two years to live without a radical improvement in his condition.

"I'm frustrated," he said. "It's either pay all this money for something that might not even work, or just live out my days. I don't want to die."