CTAD 2022 – lecanemab data hold up, with a big but
Keenly awaited results reveal no smoking gun, but lecanemab’s modest activity appears not that different from Aduhelm’s.
The numbers behind groundbreaking topline data Eisai had revealed with its Alzheimer’s project lecanemab in the pivotal Clarity-AD trial might shock the most bullish amyloid-beta believers. At worst they show lecanemab to be no better than Biogen’s controversially approved Aduhelm, albeit with less of the Aria-E side effect, the CTAD conference heard last night.
That said, beyond a perverse effect in ApoE4 homozygotes, and removal of the word “modestly” to describe efficacy in an NEJM manuscript, on the strength of Clarity-AD lecanemab looks approvable. But since Aduhelm’s main criticism is that its dataset includes a strongly negative pivotal trial, plus ineffective doses, did Eisai just get lucky by running a single study testing a single lecanemab dose?
This might be a slightly unfair criticism, since the design of Clarity-AD was driven by phase 2 data backing the single dose tested, 10mg/kg biweekly. But the study authors’ words in a draft of an NEJM paper appeared to offer a wake-up call: lecanemab “resulted in modestly less decline” than placebo, they had written.
However, in the final version of the NEJM paper, which accompanied the full Clarity-AD presentation at CTAD, that wording had been changed to say lecanemab resulted in "moderately" less decline than placebo.
1.21-point decline
None of this is at odds with what Eisai toplined in September. On Clarity-AD’s primary endpoint, effect on CRD-SB at 18 months, lecanemab was known to have performed 27% better than placebo. The full data quantify this as being a 1.21-point decline, versus 1.66 points for placebo.
In comparison, the Emerge trial cited in Aduhelm’s label shows the Biogen drug being associated with a 1.35-point decline, versus 1.74 for placebo. This makes the two amyloid-beta MAbs look very similar, as do cognitive and functional endpoints, with the caveat that lecanemab’s data are prospective whereas Aduhelm’s were cherrypicked from one study whose twin was disastrously negative.
The cross-trial comparison does reveal one obvious difference: on the CRD-SB measure, Clarity-AD patients might have been more impaired than those in Emerge. Higher scores indicate greater impairment, and the corresponding numbers at baseline are 3.17 for Clarity-AD patients on active treatment, and 2.51 for Emerge.
| Amyoid-beta MAbs in Alzheimer's disease: pivotal data across two trials | ||||||
|---|---|---|---|---|---|---|
| CDR-SB | Adas-Cog | ADCS MCI-MDL | Aria-E | Aria-E in ApoE4 | Aria-E in non-ApoE4 | |
| Lecanemab (Eisai) | ||||||
| Clarity-AD | Baseline 3.17 / 3.22 | Baseline 24.5 / 24.4 | Baseline 41.2 / 40.9 | 12.6% | 15.8% | 5.4% |
| Decline 1.21 vs 1.66 (-0.45, or -27%) | Decline 4.1 vs 5.6 (-1.4, or -26%) | Decline 3.5 vs 5.5 (-2.0, or -36%) | ||||
| Aduhelm (Biogen) | ||||||
| Emerge (high dose only) | Baseline 2.51 / 2.47 | Baseline 22.2 / 21.9 | Baseline 42.5 / 42.6 | 34.8% | 43.1% | 29.8% |
| Decline 1.35 vs 1.74 (-0.39, or -22%) | Decline 3.8 vs 5.2 (-1.4, or -27%) | Decline 2.5 vs 4.3 (-2.5, or -40%) | ||||
| Notes: all data at 78 weeks; Adas-Cog13 values for Aduhelm, and Adas-Cog14 for lecanemab. Source: company disclosures & NEJM. | ||||||
The NEJM paper reveals that Clarity-AD efficacy results were generated using a modified intent-to-treat analysis. This excluded 3% of the randomised population – 39 patients in the lecanemab cohort and 22 on placebo, who had not received at least one dose and had not undergone one CDR-SB measurement.
Yale School of Medicine’s Dr Christopher van Dyck, who presented the efficacy data at CTAD, called this a potential limitation, but said a sensitivity analysis yielded similar results.
He also tackled a separate potential concern, namely that occurrence of Aria-E, a side effect that possibly correlates with efficacy, might have unblinded Clarity-AD. Sensitivity analyses were performed, he said, and clinical raters were kept in the dark about safety assessments to minimise unblinding.
The previously disclosed 12.5% rate of Aria-E – lower than Aduhelm’s 35% – is now confirmed to have been driven by carriers of the ApoE4 allele, just as it was in Emerge.
However, the efficacy data threw up a paradox when cut by patients’ ApoE4 status. It was thought that ApoE4 carriers might derive the greatest benefit from amyloid-beta blockade, and indeed Clarity-AD had focused especially on recruiting them: ultimately 69% of the Clarity-AD population were carriers, versus just 30% in the earlier phase 2.
But this did not in fact correlate with efficacy, or at least not entirely. Dr van Dyck said there was no real difference in the effect on CDR-SB between ApoE4 carriers and non-carriers, but there was one “exception”: ApoE4 homozogotes, who carry both copies of the allele and made up 15% of the Clarity-AD population, actually did worse on lecanemab, with these patients declining 22% faster on CDR-SB versus homozygotes receiving placebo.
Just as confusingly, this numerical quirk disappeared when looking at the Adas-Cog and ADCS MCI-ADL cognitive and functional endpoints. “I don’t think we should make too much of this,” said Dr van Dyck. He suggested that this could be a chance finding in a “small subgroup” of homozygotes, who showed unexpectedly slow decline on CDR-SB in the placebo group.
Source: CTAD
Stratifying the results by disease severity showed a similar effect in patients with mild cognitive impairment versus mild Alzheimer's; conventional thinking is that patients who have milder disease will experience greater slowing of cognitive decline.
Another interesting point is that no deaths in Clarity-AD – there were six on lecanemab and seven on placebo – were “considered by the investigators to be related to lecanemab or occurred with Aria”. This will come as some relief given this week’s article in Science describing two unpublished case reports of patients who died while on lecanemab.
All that said, the authors say Clarity-AD had been designed on what looks like a conservative basis, namely that a 0.37-point reduction in CDR-SB corresponds to 25% less decline in cognitive function. As was already known, the results cleared this benchmark – in Clarity-AD and in Emerge alike.
What’s next?
Eisai and Biogen have joint rights to Adhuelm and lecanemab alike under a 2014 Alzheimer’s tie-up, but this deal was recently restructured to give Aduhelm mostly back to Biogen while the Japanese group doubled down on lecanemab.
As a result the lecanemab regulatory strategy is entirely in Eisai’s hands; the project has a January Pdufa date for an accelerated approval, but since Clarity-AD has now yielded strongly supportive controlled results a full US green light is possible, though how quickly the FDA can review the data means the timeline is uncertain.
Renegotiation of the deal has raised investor concerns that Biogen’s input into lecanemab is now minimal, with some suggesting that the group had seen little beyond topline Clarity-AD data before CTAD, though Biogen insisted at its third-quarter earnings call that the two companies were “working very closely together”.
Yesterday it emerged that the deal might have soured further. Mizuho’s Salim Sayed, who covers Biogen, said Eisai’s legal counsel had instructed the Japanese group not to respond directly to inquiries from sellsiders covering Biogen. “We previously have been able to go to Eisai with our questions regarding lecanemab,” he wrote to clients.
Either way, even if lecanemab’s absolute effect does look modest, few suggest that the dataset is insufficient to back approval. Speaking at CTAD Michael Irizarry, Eisai’s vice-president of Alzheimer’s clinical research, said Clarity-AD had been designed as the definitive phase 3 study to confirm efficacy, according to CDR-SB, which he called the gold standard of clinical assessment.
But there is a reason why companies targeting large populations used to have to carry out two large pivotal studies to support approval. It can only be guessed what data lecanemab would have yielded in a second iteration of Clarity-AD.
This is an updated version of a story published earlier.
