Terns Pharma: Derisked Allosteric Inhibitors $tern

Merck, Peak Sales, Schedule 14D-9s

Two weeks ago Merck announced an agreement to acquire Terns Pharmaceuticals for $5.7 billion ($6.7 inclusive of Tern’s cash). Terns went public in 2021, and back then was primarily focused on a treatment for MASH, a type of fatty liver disease. In 2022, Terns announced it had an allosteric inhibitor for the treatment of chronic myeloid leukemia in development (TERN-701), and that has since become the company’s lead asset. It’s also why Merck is acquiring the business.

Terns’ general business strategy has been the following: develop small molecule drugs targeting an already validated mechanism of action, but where there’s an opportunity to improve efficacy/safety/tolerability.¹

Its lead candidate is a good illustration of how this works and why this strategy is an appealing one. CML is caused by a chromosomal translocation, where a piece of chromosome 22 breaks off, merges with chromosome 9, and vice versa. Consequently, chromosome 9 is longer than usual, and chromosome 22 is shorter. Chromosome 22 now contains a BCR-ABL1 fusion gene, and in this state is described as the Philadelphia chromosome.²

This BCR-ABL1 fusion gene encodes for an abnormal tyrosine kinase that remains stuck in the ‘on’ position. Much like Ras proteins, healthy tyrosine kinases play a vital role in regular cellular processes. Critically, these kinases are meant to switch between active and inactive. In the case of BCR-ABL1, this switch between on and off doesn’t happen, and tyrosine residues are instead phosphorylated constantly/uncontrolled cellular proliferation occurs.

The good news with tyrosine kinases is they have well-defined active binding sites, and so historically weren’t considered undruggable. Kinases are enzymes that closely match the enzymes one learns about in high school biology. Designing an inhibitor, then, is just a question of designing a molecule that can inhibit that enzyme-substrate binding. This relative straightforwardness has resulted in over 50 FDA approved tyrosine kinases inhibitors for the treatment of various cancers. In the case of CML, the first approved inhibitor was Imatinib (Gleevac) in 2001. There have been a number of ‘second-generation’ inhibitors approved since then (Dasatinib, Nilotinib, and Bosutinib) with pretty direct trade-offs of increased potency for greater side-effects. It’s worth emphasizing how much of a step forward Imatinib was: this inhibitor altered CML into a cancer with a ~30% 5-year survival rate into one with a survival rate of 70%+.³

Unfortunately, these well-defined binding sites (or orthosteric sites) come with downsides. The primary one is there’s not much difference between the BCR-ABL1 binding site and the binding sites of other, non-mutant, tyrosine kinases present in the cell. The result is that healthy tyrosine kinases are inhibited, causing unpleasant side-effects. This led to a fairly fragmented market for second-gen TKIs. They collectively comprise over half of the 1L CML market, but there wasn’t a clear winner. Dasatinib and Nilotinib both peaked at ~2 billion ish in sales, with Bosutnib at a more modest ~600mm.⁴

This drastic improvement in survival rates, combined with non-ideal tolerability profiles, makes CML quite an interesting disease to develop new treatments for. The side-effects of these second-generation inhibitors are not insignificant, which leads to a high rate of patients switching medications. Terns’ management estimates that “40% of CML patients on these active site TKIs switch therapy within 5 years due to lack of sufficient treatment response or side effects.”⁵ The company also estimates that, within G7 nations, there are 17,000 new frontline CML patients every year and 13,000 CML patients who switch medications.⁶ If you can develop an effective medication with a real improvement in toxicity you can help patients for far longer than a new therapeutic in, say, the pancreatic cancer space might be able to.⁷

One solution to the tolerability issues present in 2nd gen TKIs is to alter BCR-ABL1’s activity by binding to an allosteric site rather than an orthosteric one. In the context of an enzyme, an allosteric site is a binding site that, while separate from the active site, has a direct effect on the active site. In other words, binding to the allosteric site can alter the shape of the active site, and thus increase/decrease the likelihood of substrate binding. Allosteric inhibitors can almost be thought of as putting a boot on a car. It doesn’t stop the engine from working, but the result is still a nonfunctional vehicle. Allosteric inhibitors don’t directly block the active site, but they still render the enzyme inactive.

Historically, the challenge with allosteric sites is that they’re often hard to actually locate on a protein (the reasons why are likely a subject for a different piece). By hard, I mean hard enough that D.E. Shaw’s founder has spent enormous sums of capital building supercomputers to run molecular dynamics simulations in the hope of finding these targets. Targeting allosteric sites also, however, comes with very real benefits. Most notably, allosteric sites differ between different tyrosine kinases much more so than orthosteric sites do. The result is that off-target effects should be much less, which should mean one can take dosing higher and thus more effectively attack the cancer. Targeting an allosteric site is especially helpful in cases of tumor resistance, where the BCR-ABL1 active site may have mutated so as to avoid these first-gen/second-gen inhibitors.

Terns is leveraging this allosteric approach, following in the footsteps of Novartis’ Asciminib (Scemblix), considered a third generation TKI. Asciminib, like TERN-701, is a STAMP inhibitor (specifically targeting the ABL myristoyl pocket). This ABL myristoyl pocket is an allosteric site that can lock BCR-ABL1 in its inactive position. Asciminib received accelerated approval for the treatment of third line CML in October ’21, and then received accelerated approval for first-line in October ’24. The drug has been enormously successful thus far, doing 1.3 billion in ’25 revenue with a ~4 billion peak sales estimate. In the U.S., Novartis estimates Asciminib has 23% front-line share, 57% second-line share, and 59% third-line share.⁸

Despite Asciminib’s success, there does seem to be a clear opportunity for improvement. Asciminib is a therapy that primarily competes with second-generation inhibitors on tolerability rather than efficacy.⁹ That’s valuable, but not as valuable as a therapeutic that can deliver better efficacy along with improved tolerability. Asciminib also annoyingly requires that patients fast for two hours before taking the medication and for one hour after.

The key bet with Terns is that its CML candidate can deliver on all fronts. It’s also uniquely derisked. There are a number of companies pursuing allosteric inhibitors for clinically validated targets, but very often for targets where:

1. Orthosteric inhibitors are the only validated options or

2. The target is known to be a disease driver but was previously considered undruggable.

In those cases the allosteric approach has an obvious appeal in terms of avoiding off-target effects/inhibiting a protein we previously couldn’t bind to, but there’s always the chance that binding to the allosteric pocket causes some unanticipated side effects.¹⁰ Given the approval of Asciminib, there’s no risk that the allosteric binding specifically leads to side-effects that make TERN-701 untenable. Moreover, there’s precedent that TERN-701 could move from third-line all the way to first-line, and there appears to be a decent amount of physician willingness to switch from TKIs that have been on the market for much longer to newer approaches. TERN-701 can be taken with food, and is hoping to avoid the side-effects of Asciminib (most notably pancreatic toxicity and hypertension). If that ends up being the case (and assuming other, different, side-effects don’t come up), TERN-701 could end up taking Asciminib’s 4 billion opportunity and then some.

All this said, TERN-701 is relatively early in development; the company’s most recent Phase 1 clinical data update included 38 efficacy-evaluable patients. Once this cohort is split into patients in the trial because of lack of efficacy on their last TKI (20) versus lack of tolerability on their last TKI (16) the numbers get quite small.¹¹ The market’s reaction to this data has been extremely positive, I think largely because the existing body of knowledge on effective MoAs for treating CML is comprehensive enough that a large clinical trial isn’t seen as necessary to conclude improved efficacy/tolerability.¹²

There was some debate upon the merger announcement as to whether Merck had given Terns a low-ball offer. Leerink was a proponent that the price was too low, putting peak sales at ~6.2 billion and arguing the acquisition price should sit at 1.5 - 2x peak sales/another bidder should come in. The Schedule 14D-9 came out yesterday and paints a more complete picture of the acquisition process. There were six potential bidders involved in varying degrees: Party A, B, C, D, E and Merck. Party E never responded to initial outreach, Party D said it wasn’t interested, and Party A and B both said they couldn’t arrive at a valuation above Terns’ current market capitalization.¹³ Interestingly, Party C pulled out after seeing additional clinical data:

“At the meeting, Ms. Burroughs presented new CARDINAL trial data showing TERN-701’s consistent performance compared to the data presented at the ASH Annual Meeting. However, the MMR achievement rate was lower, potentially due to more patients being pre-treated with asciminib in the evaluable population.¹⁴ The MMR achievement rate stayed within Terns’ disclosed confidence interval after the ASH Annual Meeting, with no overlap with the asciminib interval. The data supported a similarly encouraging safety and tolerability profile for TERN-701 as the data from the ASH Annual Meeting. Ms. Burroughs speculated that some aspects of the updated data may have contributed to Party C’s decision to withdraw from discussions to potentially acquire Terns”¹⁵

Party C’s withdrawal illustrates an important point when determining Terns’ fair value. The risk with this asset is that its efficacy drops/side-effects increase as more patients are evaluated, such that it ends up trending closer to Asciniminb. Such a drop would lead to a more modest peak sales number, and could result in similar dynamic to 2^nd gen TKIs where both Asciminib and TERN-701 top out at ~2 billion. That’s a perfectly fine outcome, although not a 6.2 billion one.¹⁶

Disclaimer: The information in this post is not intended to be and does not constitute investment or financial advice. You should not make any decision based on the information presented without conducting independent due diligence.

1. One very successful implementation of this idea in practice is Karuna Therapeutics, which was acquired by BMS for $14 billion. Karuna licensed xanomeline from Eli Lilly, a therapeutic that had demonstrated efficacy in schizophrenia but also caused extreme nausea. The company made a fairly straightforward modification to the drug that’s now meaningfully changed the schizophrenia treatment landscape. In other words, a validated mechanism of action that had a tolerability problem.

2. CML patients with this Philadelphia chromosome are termed Ph+, and account for over 90% of cases.

3. Different sources put this number at anywhere from 70% to 90%+.

4. These off-target effects are not always bad. Part of how Dasatinib works is by purposefully inhibiting tyrosine kinases beyond BCR-ABL1. Most notably it targets the SRC family kinases, which are thought to drive some imatinib resistance.

5. From the company’s December 8 call.

6. Page 8, 2025 Annual Report.

7. One interesting thing to think through is whether this 13,000 number will drop going forward with increased uptake of Novartis’ Asciminib. The whole reason it’s seen as attractive is because of an improved side-effect profile.

8. Pg 11, ‘25 Novartis investor presentation.

9. Asciminib’s efficacy numbers were numerically better, but not to the point of statistical significance.

10. This isn’t an apples to apples comparison, but SHP2 inhibitors are a decent example here. SHP2 was known to be a cancer driver, but we couldn’t figure out how to target it. Eventually promising allosteric approaches surfaced (including a candidate from Revolution Medicines), but they turned out to have less than ideal side-effect profiles. Big pharma ended up nixing a lot of partnerships in this space.

11. There were two patients not accounted for in this split that Terns gave.

12. A bit of skepticism is warranted here. A portion of the heart issues nilotinib can cause didn’t become apparent until after 5 years of patient follow up.

13. Terns’ stock price sat at $36/37 when Parties A and B said this.

14. MMR is a commonly used metric for measuring CML progression, and stands for major molecular response. A patient has achieved MMR when BCR-ABL1 levels are no more than 0.1% of what they were when CML was first diagnosed.

15. Schedule 14D-9, pg 22.

16. Those interested in exploring the CML space further might enjoy looking at Enliven Therapeutics, a company designing an orthosteric inhibitor with improved selectivity for BCR-ABL1 over other tyrosine kinases.