Reassuring Data Boosts Our Biogen Valuation

Biogen BIIB presented additional phase 3 data for Alzheimer’s disease drug candidate aducanumab at the Clinical Trials on Alzheimer’s Disease meeting on Dec. 5, and we’ve raised our fair value estimate to $390 per share from $383 following what we view as reassuring details.

Biogen still plans to file for Food and Drug Administration approval of aducanumab in early 2020 based on full data from its two phase 3 studies: Engage and Emerge. We’ve increased our assumed probability of approval for aducanumab to 40% from 30% and now incorporate sales of nearly $3 billion (nearly $7 billion if approved) by year 10 of our model. While this update was encouraging, we’re still uncertain about whether the FDA will see the weaker Engage trial as confirming the benefit seen in the Emerge study. We continue to award Biogen a wide economic moat rating based on the stability of its multiple sclerosis franchise, diversification in neurology with Spinraza, and a solid neurology pipeline, although we highlight the high fair value uncertainty that comes from having an innovative, high-risk pipeline.

Aducanumab development was stopped in March after a futility analysis of pooled data from two trials indicated that there was less than a 20% probability of meeting the primary endpoint. However, at that time, there was already an indication that the Emerge study was trending positive and Engage was not, and Biogen had not fully realized the extent of the impact of protocol amendments on results of these trials. As disclosed in October, additional data from more patients allowed aducanumab to meet primary and secondary endpoints (reducing cognitive decline) among high-dose patients in the Emerge study. The Engage study--which enrolled earlier than Emerge and was more affected by treatment protocols used before the amendment, which pushed patients toward lower doses--still did not meet its endpoints. Analysis of the Engage study for a subset of patients who received high, uninterrupted doses of the drug (patients sometimes suspended treatment due to side effects) seemed to imply similar efficacy as Emerge, although there were no statistics around this data, and several unknowns around this analysis led to mixed response from investors and Alzheimer’s experts.

The final, full data set revealed almost identical data to the larger data set revealed in October, with a 22% slower decline in CDR-SB (disease severity metric) for high-dose patients than placebo in Emerge, and a 2% faster decline than placebo in Engage. One of the key questions going into the Dec. 5 presentation was whether the one-month difference in trial start dates for Engage and Emerge was big enough that the protocol amendment (allowing patients carrying the ApoE4 genetic risk to take the 10mg/kg high dose) could have prevented the more advanced Engage study from meeting its endpoints. Biogen disclosed that when this amendment was made in March 2017, Engage had 200 more patients enrolled than Emerge. As a result, fewer carrier patients received the high dose, resulting in 22% of patients receiving all possible 10mg/kg treatments in the Engage study, versus 29% in the Emerge study.

Perhaps most important, Biogen provided a helpful analysis of only patients enrolled after the amendment, comparing patients in high-dose and low-dose arms of both studies with placebo patients. Among these high-dose patients, 51% in Emerge and 47% in Engage received all possible 10mg/kg doses, creating less difference in total dosing. This analysis revealed a 30% reduction in decline in CDR-SB in Emerge and a 27% reduction in decline in CDR-SB in Engage for high-dose patients, which we think shows consistency between the trials and could be enough to support approval. However, with only 55% of the full 18-month data that Biogen would have had, had the trials completed, as well as a significant dose interruption that will prevent clear data on long-term benefits to these patients even with redosing, the FDA could require another trial.