Aducanumab is one of several IV monoclonal antibody infusions that have been tested for early stage Alzheimer disease (AD); the drug missed its endpoints improving the condition, and was declared dead by the company, using futility analysis. Biogen, the company, then surprisingly resurrected the project for re-interpretation and presentation to the FDA for approval. It is a commercial product.
Many pieces have been published about it, from hopeful patients and researchers, to analysts and skeptics. Dr. Jason Karlawish, professor at Penn, has written pertinent opinion pieces. Whether aducanumab is approved, his most recent opinion is a wonderfully thoughtful discussion. He writes that it’s almost a “thought experiment,” to consider the consequences of any approved agent that might partially benefit a few, with the risk of serious side effects that might worsen patient conditions, and certainly their pocketbooks.
Karlawish walks us through the idea of a drug that slows deterioration, but does not cure it, may or may not improve the individual’s quality of life, and may or may not impact the need for eventual caregiving, nursing homes, and financial planning. He touches on the implications of discussing the pertinent genetics for a family, on practical issues of how such an agent might be used, or misused, in the current US healthcare scene, and ethical issues of how an individual might stop treatment (not dwelling on a cognitive ability to do so, and invokes uncomfortable ideas about dementia patients suffering). He cites a seminal article about AD from 1976, and the NAPA deadline of 2025. AlzGadfly hopes that when you click on the link above (or below) that it works, and you read it!
Karlawish J: Aducanumab isn’t the simple solution to the complicated Alzheimer’s crisis. STAT Published online Dec 20, 2019, https://www.statnews.com/2019/12/20/aducanumab-isnt-simple-solution-complicated-alzheimers-crisis/
He doesn’t write this explicitly, but in calling attention to the small brain swellings and brain hemorrhages, he invokes the image of a mild memory patient who then has a stroke or other permanent neurological deficit caused by the drug. There is a well-described problem in this class of agents, known as ARIA (amyloid-related imaging abnormalities) in the brain. One study shows that some patients with ARIA have eventual shrinkage of the affected areas.
AlzGadfly is admittedly an AD dilettante, but a board-certified cancer chemotherapist, and wonders what would happen if aducanumab was being evaluated by the FDA Oncologic Drugs Advisory Committee (ODAC). ODAC has been approving a record number of drugs recently; not all of them are miracle cures and oncologists are skeptical about whether some approvals really help patients. Drugs with predictably severe side effects are almost a given with chemotherapy, seemingly justified by the pace of most fatal cancer. [AlzGadfly has no opinion on FDA ODAC approvals here!]
Bevacizumab is an IV monoclonal antibody that does not cure brain tumors, but in combination may prolong some lives for weeks or months; the quality of the prolonged life is always a question. “Prolonging death,” especially one with pain, suffering, and distress is a chemotherapist’s nightmare, but it happens too often in some opinions . The parallel with AD monoclonals would be interesting to explore.
Whether aducanumab is approved or not, it will be interesting to see if someone will be bold enough to try it in combination with other candidate drugs. Yes, a BACE agent was a published failure this year, with notable adverse events, and tau agents haven’t worked so far. But the history of chemo in leukemia and other cancers is that single agents might garner partial responses. It took combination trials (Burchenal 65 years ago!) to show that drug synergy can be a first step toward durable responses, and cure. [Thanks to S. Lyman PhD who sent me the link to the Karlawish opinion piece].
Karlawish’s nice thought piece touches on but does not emphasize some of the costs of an aducanumab-based strategy in terms of capacity and workforce issues. There are a fixed number of PET scanners in the whole country that could be used to identify people with amyloid on an amyloid scan. Characterizing all older people would overwhelm the PET scans that we have. Likewise as Karlawish notes lumbar punctures to obtain CSF to characterize amyloid levels can be done, but there are not enough neurologists in the country to supervise all of those LPs with our current workforce. And that is just to identify people who would be treated.
Then we have to monitor for ARIA. This requires multiple MRI scans for each person who is treated. I don’t know the schedule for this particular trial but previous studies of monoclonal antibodies for amyloid have involved monthly MRI scans for everyone on active treatment. There are a fixed number of MRI scanners in the US. If we treat a million people with ARIA and we need monthly MRI scans, we now would introduce 12 million new scans that would have to be scheduled and obtained. Furthermore, each of those scans would have to be read.
The bottom line for me is that this is a very technologically demanding protocol that does not scale well for the US setting. It obviously scales even less well for other settings like low and middle income countries.
So there are workforce / health services sorts of issues that are also important to think about in addition to the ethical sorts of issues that Dr. Karlawish emphasizes in his thought experiment.