Manufacturers need to be more open about a dangerous Alzheimer’s drug side effect

Since the FDA’s approval of lecanemab (marketed as Leqembi) and Medicare’s recent decision to cover the drug, I have met with Alzheimer’s patients and their loved ones who are anxious to know whether they may benefit from this new treatment. They come in hope that this new medication may slow the progression of a cruel memory-robbing and personality-eroding disease.

As a practicing physician who has cared for patients with Alzheimer’s disease and other dementias for more than two decades, it is important for me to present all the benefits and risks associated with any new medication fully and without bias. I want to empower my patients and their families to make the most appropriate and evidence-based decisions about their care. It is therefore concerning to me that while the modest benefit of Leqembi in slowing the progression of Alzheimer’s has been highlighted both in medical journals and the media, relatively less attention has been paid to common side effects of the drug: brain swelling and bleeding that are together called amyloid-related imaging abnormalities, or ARIA, which have been found in all three drugs designated as “breakthrough treatments” by the Food and Drug Administration: lecanemab, aducanumab, and donanemab. I am especially worried about drugmakers’ lack of transparency in fully reporting all details of clinical outcomes related to ARIA.

ARIA have been consistently observed with most drugs like Leqembi that successfully remove sticky fragments of a protein called beta-amyloid from the brain. Recognizing the potential for rare but “serious and life-threatening” complications due to ARIA, the FDA has determined that these drugs should carry a black box warning, the highest level of precaution reserved for medicines associated with major risks of harm to patients. The rates of brain swelling and/or bleeding due to ARIA in clinical trials of lecanemab, aducanumab, and donanemab ranged from 12%-35% in treated patients, compared with 3%-13% in those receiving a placebo.

Both scientific publications reporting clinical trial results as well as drug manufacturer-sponsored awareness campaigns targeting physicians tend to portray brain swelling and bleeding due to ARIA as mostly mild or even asymptomatic, typically observed during the first few weeks of treatment, and often resolving spontaneously. Moreover, the narrative goes, in the majority of patients who show mild symptoms, they simply have to go on a temporary hold until brain scans show that ARIA have resolved, at which point the medication can be safely continued.

However, this message largely ignores the worrying reality that 1-2% of patients treated with these drugs experience serious symptoms due to ARIA including headache, seizures, delirium, impaired speech, problems with vision, and muscle weakness. Alzheimer’s clinical trials typically run for 18 months — what happens to the memory and functional abilities of these patients after these trials end? Do they fare worse when compared with patients who do not experience any ARIA symptoms or to those with only mild symptoms?

The drug manufacturers that have sponsored trials of the three FDA-designated “breakthrough” drugs have not addressed these obvious questions in peer-reviewed publications. But we do know that 12 patients in the Phase-3 trial of lecanemab developed serious adverse events due to ARIA.

The only peer-reviewed scientific publication that examined a potential link between ARIA and worsening memory in these patients was written by a team of physicians in France whose hospital was one of the centers for this clinical trial. The researchers, who are not affiliated with the drug manufacturer, looked more closely at two of the 12 patients who developed ARIA during the Phase 3 trial of lecanemab. Both had mild Alzheimer’s.

One developed severe seizures that are believed to have triggered heart failure. Eleven months later, her memory score dropped by nine points on a commonly used 30-point assessment scale of memory. The other patient developed a “massive (7 cm)” bleed in the brain and, seven months after the event, experienced persistent loss of vision as well as a 12-point decrease in her memory score. This patient is now in a nursing home and unable to speak.

To put these changes in perspective, the average decline in memory scores on this scale in patients with mild Alzheimer’s is approximately one to two points per year. These cases suggest that Alzheimer’s patients with serious ARIA may experience a significant and irreversible worsening of their memory and functional abilities.

Each of the three new Alzheimer’s drugs have also been linked with patient deaths in clinical trials — four with aducanumab and three each with lecanemab and donanemab. Some of these patients may have experienced catastrophic brain bleeding related to ARIA because they were either already on blood thinners for other conditions such as to prevent stroke due to atrial fibrillation or may have received a clot-busting drug when they experienced symptoms of a stroke. So far, only one of these cases has been published in a peer-reviewed scientific journal; it was authored by independent researchers not affiliated with the drug manufacturers. In most other instances, information on these unfortunate deaths has come to light mainly through the work of investigative journalists, including at STAT, sometimes requiring the filing of a freedom of information act (FOIA) request with the FDA.

The lack of timely and transparent reporting of these cases in peer-reviewed scientific journals by the drug manufacturers precludes any opportunity to learn critical lessons from these patients and protect others in the future. Furthermore, the maker of lecanemab has refused to share any clinical trial data with other researchers according to the data-sharing statement in their publication reporting results from the Phase 3 CLARITY AD trial. This stance is entirely at odds with the International Committee of Medical Journal Editors position that “there is an ethical obligation to responsibly share data generated by interventional clinical trials because trial participants have put themselves at risk.”

When clinical trial data are not shared with the wider scientific and medical community, it is even more important that drug companies fully disclose all clinical outcomes related to side effects of the drugs they are selling to patients. Professor Rob Howard of University College, London, and I have hence made a request to the manufacturer of Leqembi to fully report clinical outcomes related to ARIA severity urgently.

Before deciding whether to offer these treatments to eligible patients, physicians should insist on full disclosure of all observed adverse events and their relationship, if any, to worsening memory impairment and functional abilities. Regulators and scientific journals must also require both timely disclosure of these results as well as sharing of individual level patient data from clinical trials with the scientific community. Not doing so is a betrayal of the hopes of patients and risks causing them enduring harm.

Madhav Thambisetty is an adjunct professor of Neurology at the Johns Hopkins University School of Medicine. He served as a member of the FDA’s Peripheral and Central Nervous System (PCNS) advisory committee that evaluated a biologics license application for aducanumab in November 2020.