Last Sunday, lots of people were gushing about Blueprint Medicines' experimental cancer drug. So why has the company's share price dropped 20%?
The answer: biotech fund managers are paying attention to a scientist who poked holes in Blueprint's data on-stage at the annual meeting of the American Association for Cancer Research, the same venue where Blueprint presented its data.
They are also warming to Blueprint's competitor, Loxo Oncology, which is developing a similar drug.
Loxo's shares are up 8% since Blueprint's data were presented.
An M.D. Anderson researcher working with Blueprint presented impressive data at AACR's meeting in Chicago.
45% of patients whose tumors tested positive for mutations in a gene called RET saw their tumor shrink significantly enough that doctors would call it a "response" based on standard criteria. Most had non-small cell lung cancer or medullary thyroid cancer.
Blueprint said it was moving forward with a phase II trial that could result in approval of the drug, BLU-667.
Canaccord Genuity analyst Arlinda Lee increased her odds that BLU-667 would be approved from 30% to 45%, and said approval could occur in 2022 instead of 2024.
Eric Schmidt at Cowen & Co. wrote in a note to investors that he was encouraged, but warned that expectations were high, and investors might be disappointed. I wrote that the results were "impressive", but worried whether they were positive enough.
Schmidt and I should have paid more attention to those nagging doubts.
The data were first unveiled at a press conference at nine in the morning local time, but were not presented to researchers
until three in the afternoon.
After that presentation, Alexander Drilon, the clinical director of early drug development at Memorial Sloan-Kettering Cancer Center and a consultant to competitor Loxo, gave an invited commentary.
"I think it's fair though to take a step back and examine the confirmed response rate," Drilon said.
The numbers Blueprint had presented included cases where doctors had not double-checked to make sure the tumors had really shrunk enough to be called responses.
Drilon subtracted those out.
This lowered the response rate in medullary thyroid patients from 40% to 24%, and in non-small cell lung cancer from 50% to 36%.
This matters because the FDA's accelerated approval pathway uses response rate as a measure of efficacy for speedy approval.
Drilon pointed out that these new numbers were not significantly better than had been seen with Cometriq, made by Exelixis, and Caprelsa, made by Sanofi, in lung and thyroid cancer.
These drugs also target RET, but they hit proteins made by other genes, too.
Drilon did say that BLU-667 looked to have much fewer side effects than these other drugs.
Drilon's slide on the "overall response rate" to BLU-667 and two already marketed drugs that hit the RET protein.
Investors, then, are worried that the BLU-667 data may not be as good as they at first appeared. They also excited for data on another RET drug from Loxo that will be presented at the annual meeting of the American Society of Clinical Oncology in June. Another question: what will he duration of response be?
That is, once tumors have been shrunk, how long will they stay that way?
In an emailed statement, Blueprint chief executive Jeff Albers said to remember that the data are early and that the company is "pioneering a new path."
He pointed out that BLU-667 showed boad clinical activity regardless of patients' tumor type or prior therapy, even in patients who started on low doses of the drug, which may have reduced its efficacy.
He expressed hope that Blueprint would be able to rapidly enroll its clinical trial testing the new medicine.