FibroGen sinks as roxadustat safety results confuse investors

FibroGen has reported top-line results from a pooled analysis of its phase 3 roxadustat safety data. Shares in FibroGen plummeted following the readout as the wording of the number-free statement and subsequent conference call raised concerns about the safety of the kidney disease drug.

Roxadustat, an inhibitor of hypoxia-inducible-factor prolyl hydroxylase activity, moved into phase 3 in part on the strength of its potential to improve on the safety of erythropoietin-stimulating agents, such as Amgen’s Epogen, making the pooled safety analysis a key event for the program. However, FibroGen’s release raised as many questions as it answered.

FibroGen reported that the phase 3 analysis found roxadustat was noninferior to epoetin alfa against a major adverse cardiac event endpoint in chronic kidney disease patients, regardless of whether they require dialysis. 

The confusion came from qualifying statements such as “we believe there is no clinically meaningful difference in risk,” which could be interpreted as suggesting roxadustat performed worse numerically against the safety endpoints than epoetin alfa. Those doubts led investors to drive FibroGen’s stock down more than 20%, although it regained some ground as the confusion began to clear.

After talking to management at FibroGen to unpack “a very complex dataset,” analysts at Jefferies think the readout is far better than investors feared initially and ultimately a positive for the biotech and its partners, AstraZeneca and Astellas.  

“There is confusion in the market currently, but this should become clear as more investors realize the data is positive and Roxa's risk/benefit is favorable, so we see the stock rebounding and moving back up over time,” the analysts wrote.

That upbeat outlook rests on a several conclusions, not all of which are immediately apparent from the statement posted by FibroGen. The analysts think the roxadustat results in non-dialysis patients are “clearly numerically better” than those achieved by epoetin alfa, and that the experimental drug achieved a numerically lower cardiac event rate in dialysis patients. 

Buoyed by those conclusions, the analysts expect FibroGen and its partners to file for approval in the U.S. and Europe, triggering the lion’s share of the $192 million in milestones the biotech is in line to receive this year. However, some observers remain concerned about the implications of the data for roxadustat’s prospects in non-dialysis-dependent patients, a key population for the drug.

“We find it very difficult to believe there was not a pre-specified NI margin established before the data analysis occurred. Our understanding is when a NI margin of 1.3 is applied to the ITT population, then roxadustat demonstrates statistically significant non-inferiority to placebo. We do admit the stat sig improvement on decline in kidney function is an encouraging finding. However, the path forward is still very unclear to us,” analysts at Piper Jaffray wrote in a note to investors.