Maryclaire Andrews describes herself as ''by nature a cynical person … if something sounds too good to be true, it certainly is.'' But the Perth woman's reflexive scepticism has met a big challenge this year.
On February 28 she flew to Boca Raton, Florida, with her long-time partner, Basil Renolds. The following day Renolds, diagnosed with Alzheimer's disease three years ago, had the first of eight treatments he would receive over two months at the Institute of Neurological Recovery.
As he was tilted backwards at a 45 degree angle, an immune system drug called etanercept was injected between the bones at the top of his spine. The rationale is that a combination of gravity and the rich blood vessel supply in the area known as Batson's plexus will carry a concentrated hit of the medicine right into the brain, evading the filters that prevent such large molecules entering brain tissue when drugs are delivered by mouth or intravenously.
''He became less anxious, more sure, able to sign his name the way he used to,'' Andrews says. The benefits persisted and on his return to Australia, Andrews says, Renolds's score on the mini mental state examination - a test used to benchmark the condition of dementia patients - had improved.
He now flies monthly to Queensland to receive top-up doses of the drug, sold under the brand name Enbrel, from a doctor understood to be the only one in Australia trained to administer the therapy by its originator, Dr Ed Tobinick. The Queensland doctor would not allow the Herald to know his identity.
But despite Renolds's experience and that of hundreds of others in his situation, the combination of drug and delivery technique, over which Tobinick holds dozens of well-guarded patents, has never been scientifically proven through the established protocols of clinical trials in which the drug and a placebo are compared in carefully selected patients.
Tobinick's work is not without a scientific track record, but he is the author of most papers about it. His most recent publication, in the journal Current Alzheimer Research, seeks to explain the central puzzle: how the technique can achieve such dramatic results so quickly.
The drug - which targets tumour necrosis factor (TNF), a cause of inflammation - diffused into the cerebro-spinal fluid if the two were maintained in proximity by the effect of gravity, according to the paper.
This fluid, Tobinick wrote, was in direct contact with some brain cells, and it appeared these could respond to the drug without it having to pass through the blood vessels of the brain - in opposition to medical dogma.
His paper cited other scientists' findings that TNF levels were abnormally high in the spinal fluid of Alzheimer's patients, but not in their blood, arguing that the drug's direct administration into the affected region was the most likely explanation for the sudden recovery.
Ian Clark, from the Research School of Biology at the Australian National University, where he studies the inflammation factors targeted by etanercept, said Dr Tobinick was ''doing exactly the right thing. He's worked out a way to get [the drug] in that's novel and perfectly legitimate … it's not remotely a crackpot thing to do.'' Why, then, would the companies that market etanercept not want to test it extensively in their own clinical trials - the usual way new treatments are brought into the mainstream?
Enbrel, created by the US biotechnology company Amgen and marketed internationally by Pfizer for conditions including rheumatoid arthritis and psoriasis, already earns the companies $3 billion a year. An extension into Alzheimer's before an anticipated tsunami of new cases as the population ages could be a much bigger bonanza.
In Australia alone, the number of people suffering the currently irreversible disease is expected to double to more than half a million by 2030, according to projections by Deloitte Access Economics.
Tobinick told the Herald he believed the pharmaceutical industry was wary of the unusual and invasive route of administration that his technique entails.
''It's much easier to [deliver drugs] intravenously, subcutaneously or orally,'' he says.
''Even epidural is a more standard method of delivery. But none of those have the same effect.''
Pfizer's only investigation of Enbrel in Alzheimer's is a study through the University of Southampton in Britain, in which the drug is injected under the skin. Tobinick said it was doomed to fail.
''Only about half of 1 per cent of these large molecules will penetrate'' into the brain if the drug is administered that way, he says.
As well, Tobinick says, companies were heavily committed to a different therapeutic approach - developing medicines to remove the sticky plaques of the protein beta-amyloid that clog the brains of people with Alzheimer's - to divert to an older drug with less future patent protection that might nullify their investment.
Tobinick says he would welcome an industry-sponsored investigation of his techniques, and in the meantime hoped to conduct a study in primates to demonstrate whether and how Enbrel infiltrated the brain. He has already conducted a similar experiment in rats, radio-labelling the drug so it shows up using positron electron tomography scans.
However, Henry Brodaty rejects Tobinick's assessment of the lack of formal research, calling it a ''conspiracy theory''.
The Scientia Professor of Ageing and Mental Health at the Dementia Collaborative Research Centre, based at the University of NSW, says the industry has been open to trying new ways of giving Alzheimer's drugs - as witnessed by a successful small scale trial announced last week of Gammagard, derived from human blood by the US company Baxter International, which appeared to halt progression of the disease. Gammagard must be infused intravenously.
''I think drug companies are into making money,'' Brodaty says. ''If they thought [perispinal Enbrel] was a winner, they would back it.''
Just last week, Pfizer and Johnson & Johnson pulled the plug on a trial of a drug called bapineuzumab, an antibody treatment that attempted to dissolve the protein blockages in the brain.
The trial, in patients with mild to moderate forms of the disease, found no benefit to those taking the treatment compared with a placebo. It leaves only one beta-amyloid drug contender - Eli Lilly's solanezumab - in active late-stage study.
Sally Chege, a healthcare neurology analyst at the US market research firm GlobalData, said the failed bapineuzumab trial, which follows other unsuccessful beta-amyloid studies, was bad news for drug companies, which were ''redirect[ing] resources towards [cancer treatment] where endpoints are clearer and drug approval is less difficult''.
It is also, clearly, bad news for patients, who are out of options. Existing treatments Reminyl, Aricept and Ebixa improve some symptoms in a minority of patients but do not slow the underlying progression of the disease. Nothing effectively intervenes in the devastating one-way journey of Alzheimer's, which robs not only memory but ordinary daily functioning and personality, and is invariably fatal if the person does not die first from another cause.
Desperate for anything that will hold back the relentless disease even temporarily, patients clutch at straws.
An editorial last week in The New England Journal of Medicine reflected on the possible reworking of yet another old drug - the cancer treatment bexarotene - as Alzheimer's therapy. Its stunning effect in clearing beta-amyloid plaques from mice genetically engineered to develop them meant doctors would inevitably field requests from patients, wrote Justin Lowenthal from the US National Institutes of Health's Department of Bioethics.
But he concluded that even though such prescription would be legal, because of the lack of human safety and effectiveness research, ''even if patients and families are willing to take the risks for the potential benefit, the physician's answer should be no''.
Henry Brodaty fears the quest for effective Alzheimer's medications will be a much longer haul. The mice treated with bexarotene, he points out, regressed a few weeks later. Besides, mice bred to mimic a human disease may not really be a good clinical model of it. ''We've had so many promising studies in these transgenic mice but Alzheimer's is a much more complex disease than just the amyloid,'' he says.
Despite the clinical trial failures so far, beta-amyloid could still be an important drug target, says Brodaty, pointing to the discovery last month of a gene mutation that slows its production and protects against Alzheimer's - even in people otherwise at high genetic risk.
''Maybe you need multiple therapies,'' Brodaty says. ''It may be that once the iceberg has risen above the water line it is too late,'' and treatment may need to be given before any symptoms appear. But that, he says, raises its own risks, if seemingly healthy people were to undergo testing and find they were on course to develop an incurable dementia.
For Basil Renolds, it is immaterial that Tobinick's treatment - which has so far cost about $12,000 - is unproven. It does not restore all his functions, Maryclaire Andrews says, but it lets him walk the dog and confidently make a cup of tea.
''If I can keep him with no deterioration, that's a plus,'' she says. ''It's not a cure by any means but if it keeps people in a more stable state, it's got to be the way to go until something better comes along. When you look at the ballooning cost of dementia … holy smoke! Why do you just not look at this?''