Seeking Alpha • Sep 08
Mirati: A Potential Blockbuster Stock, Will Need Patience
Summary
Mirati has strong competitive differences against larger rival Amgen.
However, lead indication, 2nd line NSCLC, is not where these differences will come out.
Investors need to be patient and wait for combo data, 1st line NSCLC and especially CRC for Mirati to really stand out.
I have been regularly covering Mirati Therapeutics, Inc. (MRTX) for nearly a year now. The stock has been on a steady downward movement throughout, and has only plateaued around June from the earlier spiral. Indeed, since my last cautious coverage, the stock is up 35%.
One key reason for Mirati’s price movement is Amgen’s (AMGN) recent data from lumakras, which has again showed Mirati's superior data for adagrasib. For people in the know, to cover Mirati and not to mention lumakras is ludicrous. The adagrasib-lumakras story is thoroughly intertwined. You need to understand lumakras in order to get the Mirati story.
Mirati competes with Amgen in the KRAS G12C mutant NSCLC space. In 2nd line NSCLC, Adagrasib may have a slight edge over lumakras in efficacy, however lumakras may have an edge - also slight - in safety; plus it is ahead in clinical development, having been approved on an accelerated basis, and it has a higher median duration of response to boot. Adagrasib’s slight edge lies in the slightly higher ORR, and signs of improvement in patients with brain metastases; as well as in patients with a specific co-mutation which is known to have poor prognosis. Another important point is that adagrasib works well in combination with other drugs. For example, adagrasib has very superior data in frontline NSCLC in combination with keytruda. Lumakras' data here is much poorer.
As I said about these two points before:
Another angle with adagrasib is its CNS penetration. Mice data has shown that adagrasib has a much better brain concentration than AstraZeneca’s Tagrisso. The company is enrolling a cohort of patients to assess adagrasib in NSCLC patients with a G12C mutation and active brain metastases.
Also, KRAS and STK11 co-mutations comprise approximately 30% of KRASG12C mutant NSCLC. A potentially registration enabling Phase 2 monotherapy study in 1st line NSCLC patients with STK11 co-mutation was initiated in Q1:2021. About this program, the company said:
Patients with STK11 co-mutations or TTS scores of less than 1% have a poor prognosis with standard of care therapies with objective response rates in the first-line of approximately 30% and median progression-free survival of less than 6 months. In patients with STK11 co-mutations, published real-world data suggests that the medium progression-free survival is approximately 2 months. Collectively, these 2 patient sub-groups represent greater than 40% of the total first-line KRAS G12C non-small cell lung cancer patient population. We are discussing the potential registration pathways for each of these patient populations with the FDA.
It is also the only molecule in its class with a ~24 hour half life. Lumakras (Sotorasib) has a half life of 5.5 hours.
While this was about NSCLC, where adagrasib has some competition from lumakras, I have said before that in CRC, adagrasib’s data is beyond compare. Unfortunately for Mirati investors, the company’s lead indication right now is 2nd line NSCLC. Therefore, the market may believe Mirati is playing catch up with Amgen. This isn’t true for CRC. While lumakras has had just a 7% response rate in CRC, Mirati showed a 22% response rate and a median duration of response ((MDOR)) of 4.2 months. As the company has previously guided, anything better than 20% and 4 months respectively is good enough for seeking accelerated approval. Recently at ESMO 2022, Mirati again showed its superiority both as a monotherapy and as a combo in CRC versus lumakras - see here.
In 2nd line NSCLC, though, while Amgen managed to push in a very late breaking abstract from its CODEBREAK-200 confirmatory trial, the data readout pushed up MRTX stock because it throws some doubt on lumakras’ full approval based on this data alone.