ESALY - Biogen And Eisai's Leqembi: A Dose Of Reality For Alzheimer's Disease

2023-03-16 08:45:38 ET

Summary

• More than likely the Food and Drug Administration will grant full approval to Leqembi in July which should boost Biogen and Eisai's stock value.
• If the FDA grants full approval, Medicare has indicated that it will expand coverage for the drug.
• Biogen and Eisai compared APOE4 carriers (with one and two copies of the gene) and non-carriers against the combined placebo decline rather than against the placebo decline for each group.
• Doing the latter would probably have revealed that non-carriers barely experienced any benefit from the drug.
• APOE4 carriers show some slowing in decline, but face substantial risks from Leqembi including potential brain bleeds and swelling that can result in death.

Biogen ( BIIB ) and Eisai ( ESALY ) are confident that the FDA will grant full approval for Leqembi for Alzheimer's disease in July of this year. Despite pressures from the Alzheimer's Association and from various senators and representatives, Medicare has held firm on its stance not to extend coverage to Leqembi outside of any further clinical trials. However, Medicare has indicated that it will broaden coverage if the FDA grants full approval ( article ). At this point, this seems to be the most likely scenario. And if it plays out this way, Biogen and Eisai should see substantial increases in their stock value.

However, a full examination of the data may yet result in some restrictions. There are a number of issues which Biogen and Eisai have yet to address. First, the results from the phase 2b trial are not consistent with the phase 3 trial. Most importantly in the phase 2b trial, APOE4 carriers saw a greater slowing of decline in cognition than non-carriers whereas in the phase 3 trial it was the reverse.

Phase 2b trial results ( BAN2401/lecanemab )

• APOE4 carriers 63% less decline in ADCOMS
• Non-carriers 7% less decline in ADCOMS

Phase 3 trial results ( p.33 , access limited)

• APOE4 carriers 8% less decline in CDR-SB
• Non-carriers 41% less decline in CDR-SB

ADCOMS: Alzheimer's Disease Composite Score

CDR-SB: Clinical Dementia Rating-Sum of Boxes

Part of the discrepancy is due to the fact that the phase 2b trial compared APOE4 carriers and non-carriers on lecanemab against the mean placebo decline for each group whereas the phase 3 trial compared APOE4 homozygotes (two copies of the gene), APOE heterozygotes (one copy of the gene), and non-carriers on lecanemab against a combined mean. How this skewed the results for the phase 3 trial will be examined below.

In the phase 2b trial, Biogen and Eisai used ADCOMS as their primary endpoint whereas in their phase 3 trial they used CDR-SB. This makes direct comparisons between the two trials difficult but not impossible.

Phase 2b ADCOMS change from baseline at 18 months ( pp. D46, D47 )

Placebo Change 10mg/kg monthly 10mg/kg biweekly

• All Groups .172 .142 .126
• APOE4 carriers .180 .139 .096
• Non-carriers .146 .143 .135

In its phase 3, Biogen and Eisai only provided a mean placebo decline for ADCOMS: around 2.2 points ( from graph, p. 28 ), which is 1.28 times higher than the placebo decline in the phase 2b trial (due to slightly higher baseline). Assuming the same rate of decline for all groups (which may not be the case) and considering that CDR-SB scores are about 7.5 times greater than ADCOMS scores, the CDR-SB scores for the phase 3 clinical trial would be the following (in the phase 3 trial only the highest dose was used).

Phase 3 calculated CDR-SB scores change from baseline at 18 months

Placebo Change 10mg/kg biweekly

• All Groups 1.66 1.21
• APOE4 carriers 1.73 .92
• Non-carriers 1.40 1.30

For CDR-SB scores, Biogen and Eisai reported that APOE4 non-carriers declined by .75 points less and APOE4 heterozygotes declined by .5 points less on lecanemab than those on placebo. But APOE4 homozygotes baffling worsened by .28 points versus placebo.

If APOE4 homozygotes actually got worse on lecanemab, the simple solution would be to reject the drug for APOE4 homozygotes, who are the most likely to suffer severe side effects from the drug.

But the complicating factor is that APOE4 carriers decline more rapidly during the early stages of Alzheimer's disease, with homozygotes declining more rapidly than heterozygotes ( article ). For instance, if APOE4 carrier homozygotes declined from the placebo baseline of 3.17 to 5.6 points (which is within the realm of possibility ) over 18 months then APOE4 homozygotes would have "improved" by .49 points on lecanemab versus placebo not worsened by .28 points.

• 5.6 - .49=5.11
• 4.83 (3.17 + 1.66) + 2.8=5.11

Few adjustments need to be made for APOE4 heterozygotes as they are not far from the combined mean. Once homozygotes are analyzed separately, the mean placebo decline for the heterozygote group drops a bit while the mean for the highest dose group may rise a little with larger numbers of participants taking the 10mg/kg biweekly dose in the phase 3 trial ( snafu ). The .5 reported slowing down in decline for APOE4 heterozygotes is probably close to accurate.

The one set of calculations that does not work well between the two trials is for non-carriers.

When compared against the combined mean, non-carriers on lecanemab declined by .36 points less (based on phase 2b trial results):

• 4.83 - .36=4.47 (1.66 - 1.30=.36)

When compared against the mean for the non-carrier group, non-carriers on lecanemab declined by .1 point less (based on phase 2b trial results).

• 4.57 - .1=4.47 (1.40 - 1.30=.1)

When compared against the combined mean, the non-carrier drug group in the phase 3 clinical trial declined by .75.

• 4.83 - .75=4.08

However, the mean placebo decline for non-carriers may also have been less due to the higher baseline in the phase 3 trial (3.17 instead of 2.9):

At best, non-carriers declined by .49 points less rather than .75 points less after using the non-carrier placebo mean rather than the combined mean (1.66-1.40=.26) and probably much closer to that of the phase 2b trial (a .1 point slowing of the disease). This would be consistent with the results from every other current anti-amyloid drug trial ( Aducanumab/Aduhelm p.58/12 , donanemab , tramiprosate/ALZ801 ).

APOE4 carriers have more amyloid in their brains than non-carriers which in part explains their more rapid progression at least during the early stages of Alzheimer's disease than non-carriers. Removing amyloid brings their rate of progression closer to non-carriers. For non-carriers amyloid appears to play only a small role in disease progression and removing it has very little impact. And thus is the limitation of every anti-amyloid drug including Leqembi.

A second problem for Leqembi is that it was only tested on patients with mild cognitive impairment and very mild Alzheimer's disease patients. It is not clear if even the modest slowing down the disease in APOE4 carriers would occur in those with "solidly" mild Alzheimer's disease. More amyloid oligomers (which Leqembi) removes are converted into less-toxic (or possibly non-toxic) plaques such that the removal of amyloid oligomers has less and less impact as the disease progresses.

Lastly, lecanemab/Leqembi has likely led to at least three deaths in patients. Two of the patients were also taking blood thinners. Biogen and Eisai has not acknowledged that the drug was responsible for the deaths, but brain bleeding and brain swelling are known side effects in APOE4 carriers. For some people at least, the treatment may be worse than the disease, which is a lot to say when one is talking about Alzheimer's disease.

In summary, there is a modest slowing down of the disease in APOE4 carriers, but that slowing down may not be clinically significant and comes with potentially fatal side effects. There is very little slowing down of the disease in non-carriers.

The FDA will probably not consider any of this when making its decision, even though it should. Biogen and Eisai should experience a substantial boost if it gets the drug over the finish line, but other than stockholders there will be few winners from this outcome.

For further details see: