Kishore S. writes: "Good article and you seem to have uncovered some points/facts that the whole world missed out on including Baker Brothers. Is it possible that everyone was so blind to the obvious that small investors like me have to get burnt so badly? Anyway, good job on adding up the numbers where all the other experts are blind." The lesson here is know what you own. Do your own homework. Don't just buy a stock because a big hedge fund owns it. Avery_t writes: Even if tasilmelteon doesn't work for its initial target population or end goal, it might still be a great sleep agent for the general population of insomniacs, who have very, very few choices. Any compound that induces sleep without A) being addictive or B) lowering the quality of sleep is a potential gold mine. It could be a paint thinner that's discovered to work as a sleep agent. It doesn't matter. Takeda tried with this approach with Rozerem but failed badly. Rozerem (scientific name: ramelteon) is very similar structurally to tasimelteon. Both drugs are melatonin agonists that hit the MT1 and MT2 receptors in the brain. (Unlike other sleep drugs, like Ambien, which works against GABA receptors.) Rozerem was approved as an insomnia drug in 2005, which led to this memorable commercial featuring Abe Lincoln and a talking beaver. Great ad. Unfortunately, Rozerem is a mediocre prescription sleep aid. You have to go back a bunch of years, but Vanda's original idea was to develop tasimelteon as a sleeping pill. Phase III studies were run which produced okay results -- nothing spectacular. But then Rozerem was approved and flopped commercially, forcing Vanda to go disease-shopping for another way to develop tasimelteon. That's how the idea of using the drug to treat non-24 was hatched. Tasimelteon has no alternative future as a regular sleeping pill. Lazard Capital analyst Josh Schimmer is the sell-side's most vocal Vanda table pounder. On Wednesday, citing his "access to management," Schimmer came to tasimelteon's defense: The FDA worked with VNDA during SET/RESET to optimize endpoint selection. As such, we have little concern that the endpoints shifted over time, and the company was not changing endpoints surreptitiously in response to blinded or unblinded data. Phase 2 results revealed too much noise exists in N24 to use TST as an endpoint, so the FDA recommended more suitable assessments. The ultimate choice of endpoints was finalized in conjunction with the FDA before the data set was locked. The primary endpoint of entrainment was selected as a true measure of the disease's root cause and is strongly supported by KOLs. As a biomarker, however, it was important that secondary endpoints which capture the effects of tasimelteon were positive, which they were. The N24 Clinical Response Scale was created by the company, but validated in the SET/RESET studies by key secondary endpoints. Furthermore, the FDA agreed in the pre-NDA meeting that the data were adequate to support approval. Schimmer engages in some wishful thinking. The FDA told Vanda during a pre-NDA meeting that the tasimelteon data "were adequate to support approval." Really? That's a remarkable statement for Vanda to make because it re-writes decades of regulatory regulations and precedent. Why spend 10 months reviewing a drug for safety and efficacy when it can be done over the course of a 60-minute pre-NDA meeting? I'd like to hear more from Schimmer or Vanda explaining precisely how the clinical response scale -- made up by Vanda during the last months of the tasimelteon trial -- was "validated" by other endpoints in the same trial.