Skip to main content

Clinical Trials

Tetraphase and Their Troubles

I wanted to provide an update of sorts to a piece from a few years back. This was the lawsuit brought by Mark Charest against Harvard et al. over royalties relating to Tetraphase, the antibiotics company founded by Andy Myers of their chemistry department out of his group’s synthetic work in the area. Charest believed that he had been shortchanged on potential royalties, as that earlier blog post details, but he and Harvard eventually reached an out-of-court settlement.

It would be interesting to know if that was a cash transaction, or if Charest just got the royalty rates he was seeking. That’s because the company’s lead antibiotic, eravacycline, has been having a rough time of it. The compound failed a clinical trial in 2015, and a few days ago, news came that it had missed endpoints in another, both of them in complicated urinary tract infections. Eravacycline was well tolerated, but the trial was trying to show non-inferiority to another antibiotic entirely (Merck’s carbepenem compound ertapenem), and it did not. It also missed another primary endpoint in response rate data. (The previous urinary infection trial was aimed at non-inferiority to levofloxacin).

Eravacycline is a fluorinated tetracycline, of a sort that would be probably very difficult to come by other than through total synthesis. It has demonstrated non-inferiority (versus meropenem) in complicated intra-abdominal infections, and the company has submitted an NDA for that indication. But overall, it has not been easy for Tetraphase, as the stock price demonstrates: back just before the 2015 failure, TTPH hit $50/share, and it’s currently just over $2. This in a biopharma bull market.

This history illustrates several things about antibiotic development. You may, for example, be wondering about all this non-inferiority stuff. That’s a route to approval because of drug resistance – it’s expected, naturally enough, that antibiotics of different compound/mechanistic classes will be able to pick up the slack for each other as resistant organisms develop, so if you can show that a new agent is at least not inferior to an existing one of another class, then the drug is presumed to have value – soon enough, anyway.

There’s a perception, among people who don’t actually know what they’re talking about, that antibiotic development must not be all that bad. After all, you have direct assays to kill the exact organisms that cause infections (and their variants), and you can see just how potent various compounds are in the most direct way possible. The animal models aren’t bad, either – if you can clear infection in a small mammal, your chances of doing so in a larger one are pretty decent. The connection with human disease is (in many cases) admirably close. The problem is finding compounds that will actually do anything useful in these assays. Having tried some of that myself, I can testify that it is quite painful (and believe me, I didn’t get to experience all the field has to offer in that line).

Good bacterial targets are not all that easy to come by, and when you do find them, screening them against (even very large) compound collections often yields nothing actionable. And if your assay is against the protein target alone, you may well just be fooling yourself, because actually finding active compounds that will get into whole living bacterial cells in useful amounts is extraordinarily frustrating. Bacteria, after all these years of fighting it out, have evolved layer after layer of defense mechanisms to keep that very thing from happening, and they’re capable little creatures. Even if you do take them unawares, it won’t be for long – their rapid generation times (and their annoying tendency to actually speed up their mutation rates under stress) mean that they’ll probably find a way out of your trap pretty soon.

So I wish Tetraphase well. We most definitely need new antibiotics, and as an organic chemist I’m glad to see total synthesis potentially riding to the rescue. But it’s a hard business!

27 comments on “Tetraphase and Their Troubles”

  1. Synthon says:

    Bacteria are capable little creatures. 🤣

  2. Mister B. says:

    After a look on the Wiki page of the listed compounds. Quick question.

    Is there a room for small molecules in antibiotics ? Or are we definitely good to macro-compounds, peptides, ADc ?

  3. Anon says:

    @ Yes, there is room for small molecules in antibiotics. The problem is that of resistance and people are running out of basic frame and am referring to beta-lactams, tetracycline etc. And, first 50 ideas we can think off – its all done 40 or 50 years back! Macrolides (novel) and the semi synthetic version may offer some novelty that one can easily patent. It is literally needle in haystack!

  4. Bagger Vance says:

    I seem to recall folks thought Charest would “never work in this town again” after going against his old boss, but a quick look online dispels this notion. He appears to have been constantly employed and moved on to investment banking ($$$$$).

    1. Me says:

      If everyone who fell out with their academic supervisor never worked in the field again, there wouldn’t be any scientists!!

    2. Hap says:

      I thought he was already in investment banking when he filed the suit.

      I imagine Myers has pull in pharma and fine chemistry (because he can hinder or facilitate access to his students and expertise), but outside of that, I wouldn’t have assumed that he would have had enough pull to prevent Charest from getting a job in other fields.

      1. Bagger Vance says:

        Hap: I think you’re right, but I also think the Harvard name opens doors at investment/finance/consulting, but if those folks want Letters of Rec (probably more than publications)…?

        Me: lol, true enough, i’m at least there with one former advisor

        1. Hap says:

          One of his complaints, I think, was that Myers would not recommend him. It may have hurt getting other jobs and it would have been hard to explain around. I assumed that the lack of recommendation led him to decide he had nothing to lose in suing for what he considered a fair share of the patent output, so it had to have been significant. On the other hand, he did get another job – I just don’t know how long or hard it was, or how much less he might have gotten paid than with his recommendation.

  5. Barry says:

    there are multiple modes of bacterial resistance, but exclusion or active transport of small molecules out of the cytosol is huge. That’s why targets on the bacterial outside are so important. Beta-lactams, vancomycin…

  6. Antibiotic Maker says:

    And another overlooked area of development is the actual necessity of the target in a host wound model. Many potent antibiotics that clear infections are inactive in the host as the environment, gene expression, and nutrient availability is very different. IMO combo therapies are the way going forward, a few big papers on new classes of molecules and targets are in the pipeline though. Keep an eye out…

    1. Calvin says:

      I was off that view as well, that combinations might be a good idea. However, we are simply reinventing the problems of the past where this approach did not work and caused more problems than it solved. Have a read of “The Antibiotic Era:Reform, Resistance and the Pursuit of Rational Therapeutics”. Interesting story. Having spent time working in the field, first as a med chemist (Derek is right, it’s a ferociously tough field) and then in a more policy area I came to the conclusion that the absolute best way forward is more broad(ish) spectrum small molecules. Hard to find, but absolutely the best way. For lots of reason, too detailed to debate here. But combinations (and this still surprises me) are probably not the way forward….

      1. woodstock says:

        Can you give a bit more explanation on why you think combinations aren’t the way forward with antibiotics? I originally come from the HIV world where combinations are an absolute necessity and this similar line of reasoning is true for HCV as well. What makes bacteria different in this respect?

        1. johnnyboy says:

          Am not a specialist, but my understanding is that resistance could arise as a consequence of indiscrimate use of commercial combos (already pre-mixed), where concentration of each component of the combo may not be optimal for the individual usage, and where bacteria may be sensitive to one component of the combo but resistant to the others, which may drive resistance to the other component’s class. Combo therapy makes more sense when you can control the dosage of each component separately, but pre-mixed drugs don’t offer that flexibility.

          1. MikeRobe says:

            Resistance is also triggered by common consumer chemicals unrelated to antibiotics, a secret antibiotic resistance zealots dont want you to know.

  7. hn says:

    Derek, what do you mean when you say that HTS gives few results that are actionable? What makes them not actionable?

    1. Andy II says:

      Several excellent review articles already describe the challenges in this topic well, i.e., Tommasi R, Nat Rev Drug Discov. 2015 Aug;14(8):529-42. ESKAPEing the labyrinth of antibacterial discovery. And, another problem is that even when a novel antibiotic was developed, it will be saved as a just-in-case in order to prevent the targeted bacteria from developing a resistance and therefore a company can’t make money. People are talking about creating incentives (i.e., pull incentives) for companies to work on antibiotics field, but we still don’t see reasonably attractive incentives yet.

    2. Derek Lowe says:

      Very low hit rates, chemical matter that doesn’t show a real SAR, etc.

      1. Andy II says:

        But if you work on the similar scaffold/class (i.e., fluoroquinolone, cephalosporin, penem, macrolide, aminoglycoside, tetracycline), you can identify a “new” compound easily, aka high hit rate. However, these “new” compounds would likely end up with those that had been covered by prior art/patent and most importantly suffer from drug resistance by the bacteria from the existing the same class.

      2. milkshaken says:

        my experience was radically different. In 90s, we did some pilot HTS antimicrobial growth assay on a collection of purified in-house-made building blocks that came from scaling up intermediates for combichem. It was some 100 compounds in the collection. Most of these compounds I made myself and the rest was few commercial building blocks. The purpose of the training set and its screening in few pathogens was to find out 1) if HTS can be done on growth inhibition assay, and how laborious it will be if done manually 2) what kind of hits comes out, if they have a real SAR (so there was a quick follow on). I was surprised that the assay produced real-looking hits with MIC activity in single digit microgram/mL. There were some nonspecific membrane active detergent-like molecules (known to be antimicobial and hemolytic too), and some triclosan-like looking greasy monsters, but I was pleased that two real known lead structures came out: Nitrofurfural and its derivates (approved antibiotics nitrofurantoin is its semicarbazone) , and a simple carboxyethyl- substituted beta lactam that looked quite a bit like monobactam. I was disappointed at the time the company decided they did not want new antibiotics and killed the project.

        1. anon says:

          that nitrofurantoin is a drug, let alone a WHO essential drug, is mind-boggling. It is about as ugly a structure as one gets…

          1. openmind says:

            this bias is part of the problem plaguing medicinal chemistry in drug discovery. such ingrained bias for what a molecule should look like, open scorn for new types of molecules, and then surprise then the pipeline is running dry…

          2. milkshaken says:

            there are few nitrofurfural hydrazone antibiotics used in humans, there is also a group of nitroimidazole antibiotics – rather blunt tools, but quite helpful if you get some nasty tropical disease

    3. ScientistSailor says:

      See Also:
      Drugs for bad bugs: confronting the challenges of antibacterial discovery
      David J. Payne, Michael N. Gwynn, David J. Holmes & David L. Pompliano
      Nature Reviews Drug Discovery volume 6, pages 29–40 (2007)

      In short 75 HTS campaigns over ~10 years, maybe one project moved into Phase 1.

  8. Thomas Lumley says:


    In a lot of bacterial infections, resistance is primarily problem if already-resistant bacteria cause the infection. If you can get high enough concentrations of an effective antibiotic agent to the bacteria, you can get a cure, and there isn’t a big advantage to adding another agent. The main reason for combinations is empiric therapy for urgent care, when you don’t have time to find which single agent will work. Even then, you might often send a sample to the lab, start empiric treatment, and then switch to a single agent when the lab results come back.

    In HIV (don’t know about HCV) and in some bacterial infections (such as tuberculosis), resistance developing during the course of treatment is a big deal. Combination therapy is then absolutely critical to make sure that the microbe can’t mutate its way around the treatment.

  9. Mach4 says:

    Eravacycline while an interesting synthetically accessible drug looks too much like Tygacil at the C9 position and probably is in a Wyeth/American Home Products/Pfizer Markush structure and patent somewhere. So if its not the SEC or other issues dogging Tetraphase its the relative activity or inactivity of their flagship drug hurting their efforts. They had thousands to choose from and they chose a glycylcycline analog? Ouch That hurts!

    But Andy Myers must be congratulated on the total synthesis of a tetracycline, which history teaches several other groups succeeded as well- all leading to failure in the clinic. And nobody cares about the Harvard-Charest dynamic- just shows the insanity and human misbehaviors surrounding gold colored compounds and how it makes mere mortals weak and depraved at all levels. And nobody cares about Myers, Pharma or anything from Industry when it comes to antibiotics- they all have left the field or pretend to be busy.

    Buts that’s the history of the tetracyclines, great drugs that have still unlocked potential against bacteria and in inflammation and neurodegeneration, for those who dare to work with them!
    Beware! They are cursed!

  10. ScientistSailor says:

    Derek, Eravacycline was well tolerated because they weren’t giving enough to be efficacious! Recall that in their oral trial the lead-in phase evaluated 200 vs 250 mg, and they went forward with 200, due to tolerability.

  11. Design Monkey says:

    That company name kind of calls up associations with Triphase thingy from movie Extreme measures. And that Triphase thingy was creepy (doing medical research Joseph Mengele style). So, nothing to do with antibiotics, but still not the best way of naming company in med research field.

Comments are closed.