There have been some pretty dramatic clinical trial results coming out recently, and unfortunately drama is a variable that can take either a positive or a negative sign in front of it. On the plus side, MacroGenix, a company that not many people had been paying attention to, announced results of a head-to-head trial of their latest HER2 antibody (Fc-optimized) against Genentech’s Herceptin in a tough patient population (third- and fourth-line treatment in HER2+ breast cancer). The company hasn’t announced many details, but they apparently reached statistical significance versus Herceptin (standard of care), and no one seems to have been expecting that. You can tell that by the way the stock reacted, more than doubling on the news.
As mentioned here, some of that move is covering of short positions, and that gives me a shiver, since I’ve been in just that position once. The flip side of “long and wrong” is “short and mort“, to dip into French for the sake of rhyme. But it’s true that not many people gave this trial a chance – that link (Evaluate Pharma) has a good point about the fact that overall survival has not read out yet and that all we have now is a p-value:
Investors keeping the faith must now believe that US filing will not depend on OS, and that Macrogenics really has discovered something big rather than simply having got ridiculously lucky. And Macrogenics must pray that the US FDA approves margetuximab before OS data read out.
For the sake of breast cancer patients looking for their third or fourth option, let’s hope that this wasn’t just a statistical fluke – but we’ll have to wait on the numbers. So far, the effect looks even better for carriers of a known Herceptin resistance mutation, which is probably a good sign. More data!
So much for the good news. Solid Biosciences unfortunately got the other end of the stick – they’re trying to develop a gene therapy for Duchenne muscular dystrophy, using a viral vector to splice in the gene for a dystrophin construct. That would, if it works, allow treated patients to produce the very thing they’re lacking, surely with major effects on the disease. But it’s been hard – gene therapy is nowhere near being a done deal for anyone. It’s a minefield, and we don’t know enough about the delivery and expression of such gene constructs in humans to do anything more, in most cases, than discover those mines by jumping up and down on them.
That’s what happened when Solid dosed their first patient some months ago – there was serious trouble, with decreased platelets and red blood cells, and signs of a strong immune response. That led to a clinical hold from the FDA which went on for some months, but they were able to dose a few more people eventually. But of the three readouts reported (just doing Western blots for the new protein off biopsy samples), two have undetectable levels of dystrophin, and the third was well below the levels needed for therapeutic effects. That’s the thing: given our current knowledge of transcription in humans, we don’t really know how to avoid results like this. You just have to load up your virus (or whatever), dose your gene, and see what happens, and sometimes this is exactly what happens. Solid Bio’s stock immedately dropped over 70%, as well it might.
This sort of thing doesn’t just happen to small companies, though. Not by any means. Eli Lilly sells Lartruvo (olaratumab), an antibody against PDGF-alpha, for the treatment of sarcoma, dosed with doxorubicin. (Those with long biopharma memories will be interested to hear that this one was an Imclone project). Correction: Lilly used to sell that antibody for that indication until now, because the combination was approved on an accelerated/conditional basis after promising Phase II data. But the Phase III readout is here, and it’s worse than anyone expected.
Since we were talking statistical significance up there a few paragraphs ago, it’s worth mentioning that the Phase II trial for olaratumab just missed significance for progression-free survival, but was highly significant in overall survival. That’s what led to the accelerated approvals in the US and the EU, but both agencies (as usual in such cases) want to see more data. But the Phase III showed no statistical difference at all in overall survival for the antibody + doxorubicin versus doxorubicin alone. This was a larger trial (485 patients versus 133), and if your effect is real, running a larger well-controlled trial in similar patients should not cause it to disappear. I think we can all agree on that principle of clinical science. Lilly issued a downgraded forecast for its earnings, partly on the basis of these results, since Lartruvo had already brought in over $200 million in revenue last year. As far as I know, the antibody is still in trials with a different chemotherapy combination, but this is not an encouraging sign.
As usual when something like this happens, this is one to consider when thinking about accelerated approval, compassionate use, right-to-try and similar issues. I think that granting such approval was the right thing to do in this case, since the patients involved very much need a chance at better treatment. But the key word is “chance”. The Phase II data were enough to make a reasonable person think that it was a chance worth taking, but drug development is not a reasonable field of study.
There’s a common theme running through these three stories: the need for more data, and for more understanding. Why did the MacroGenix antibody work as well as it seems to have worked? Why did Solid Bio’s gene therapy fail to cause any new protein to be produced? Why did Lilly’s drug not actually do anything for sarcoma patients, even though mechanistically it really seemed as if it should? We don’t know, we don’t know, we don’t know, and what you’re seeing is how (slowly, expensively, painfully) we find the answers.
Update: and later in the day, Sangamo reported on their zinc-finger gene editing clinical results. There are headlines about historic firsts, but the bottom line appears to be that they got very low expression of their desired gene, and one patient seems to have had an immune response to the treatment. Not what they were hoping for, for sure. A later version of this technique might do the job, but this won’t. Gene-editing stocks are getting hammered across the board at the moment. Over-pessimism, or just a return to reality after too-high expectations?