EUR 850.00 For Business Accounts Only

Morningstar | Maintaining our Biogen FVE and Alzheimer's Forecast Despite Confusing Analysis for BAN2401 Trial

Eisai presented key subset analysis for the phase 2 study of Eisai and Biogen's amyloid antibody BAN2401 on Oct. 25 at the Clinical Trials on Alzheimer's Disease, or CTAD, meeting in Barcelona, and we're not making any changes to our Alzheimer's sales estimates for Biogen or our $420 Biogen fair value estimate at this time. We still assume $10 billion in amyloid antibody sales by 2027 if both amyloid antibodies are approved, using a 60% probability of approval for both drugs. Even if we remove both amyloid antibodies from our valuation, we arrive at a fair value estimate near $350, above recent trading levels. In addition, this $350 valuation factors in virtually no long-term success for the remaining early-stage Alzheimer's pipeline (including tau programs), which we think is unrealistic. We still see Biogen's steady neurology focus and increasingly diverse portfolio and pipeline supporting a wide moat. While Biogen and Eisai are still discussing the idea of a phase 3 program for BAN2401 with regulators, so there is uncertainty on the future of this program, aducanumab's phase 3 studies are fully enrolled and final data are expected in early 2020. Overall, we think data supports the conclusion that both drugs (which target aggregated forms of amyloid) reduce plaque burden and slow clinical decline at high doses, although inconsistencies in subset data for BAN2401 are enough to prevent us from raising our assumed probabilities of approval.

The updated BAN2401 carrier subset data for the high-dose arm versus placebo appeared to refute the idea that an imbalance in APOE4 carriers between the high dose arm and the placebo arm drove the initial positive results in July, which was reassuring and gave us more confidence in BAN2401's efficacy. As we discussed in our July 26 note, the phase 2 study of BAN2401 had strong overall efficacy data at the highest dose, with a significant slowing in decline as measured by ADCOMS (30% slower decline) and ADAS-Cog (47% slower decline) versus placebo, and a trend to a benefit on CDR-SB (26% slower decline). Importantly, subset analysis showed that among placebo patients, carriers and non-carriers saw declines in ADCOMS (the composite clinical endpoint used in the trial) that were similar to one another. As we previously discussed, the imbalance among APO4 genetic variants between the highest dose arm (30% of patients were carriers) and the placebo arm (70% of patients were carriers) caused confusion and debate surrounding the interpretation of the initial 18-month data.

However, that reassurance turned back to confusion, as the carrier subset clinical endpoint data suggested that APOE4 carriers could actually be responding better to treatment than noncarriers, which would be puzzling, as carrier status was not a factor for clinical responses in the aducanumab phase 1 PRIME study. In addition, high-dose BAN2401 patients in the non-carrier group appeared to do slightly worse than placebo on CDR-SB, which is the chosen primary endpoint for Biogen's phase 3 studies of aducanumab. We think the small sample size for the analysis, which was partly due to high discontinuation rates (regulatory action to remove carriers from high-dose arm of the study as a precautionary measure), and the general variability in clinical endpoint data (which is largely based on questionnaires) make it difficult to conclude that APOE4 carriers respond better to treatment. We're inclined to put more reliance on the similar reductions in plaque levels (a much more objective, quantitative measure) for carriers and noncarriers in the BAN2401 high-dose arm.