Skip to main content

From osimertinib to preemptive combinations

*March 2024*

Note to reader: Although Tagrisso is the preferred first-line treatment, some patients who took a 1st or 2nd generation drug as first-line (Tarceva, Afatinab) and then Tagrisso as second-line experienced a much longer PFS (progression-free survival) than those who started with Tagrisso.


Here, I suggest that while first-line osimertinib extends median progression-free survival (PFS) in EGFR-mutant lung cancer compared to first-generation TKIs, it reduces individual PFS in 15–20% of patients compared to first-generation TKIs. Since detecting a single resistant cell before treatment is usually impossible, osimertinib must be used in all patients as a first-line treatment, raising median PFS overall but harming some. The simplest remedy is a preemptive combination (PC) of osimertinib and gefitinib. A comprehensive PC (osimertinib, afatinib/gefitinib, and capmatinib) could dramatically increase PFS for 80% of patients compared to osimertinib alone, without harming anyone. This article also explores PCs for MET-driven lung cancer.


In EGFR-mutant-dependent non-small cell lung cancer (NSCLC), first- or second-generation EGFR-TKI (e.g., gefitinib, erlotinib, afatinib, and dacomitinib) select for resistance due to the T790M point mutation (EGFR T790M) in 50% of patients [14]. In other words, before therapy, 50% of patients have pre-existing resistant T790M mutations. Only one or a few cells contain the T790M mutation, because this resistant mutation confers no selective advantage prior to therapy. As soon as treatment starts, these rare cells selectively proliferate and eventually produce billions of cells [5], rendering the tumor resistant to the 1st/2nd generation of tyrosine kinase inhibitors (TKI). The cell with T790M is sensitive to the 3rd generation TKI osimertinib.

In 2015, the FDA approved osimertinib for NSCLC with a T790M mutation as a second-line therapy. This approval as a second-line therapy was based on the understanding that an untreated tumor cannot be T790M-positive; the mutation may initially occur in only one cell. Furthermore, 50% of patients do not harbor this mutation. It might seem logical to administer osimertinib after a tumor develops resistance to first- or second-generation TKIs, given the uncertainty regarding which patients will acquire the T790M mutation.

Consider a hypothetical scenario: If osimertinib were a combination of two medications (the first targets oncogenic EGFR without secondary T790M and the second targets only oncogenic EGFR with secondary T790M), then an oncologist would not prescribe the second drug to a patient sensitive to the first drug. And why would an oncologist prescribe the second (anti-resistant) drug? The tumor is not resistant to the first drug. Even if a patient has a single cell with T790M, its killing will be unnoticed at first, even in the months. The response rate will not be affected too. But after a year, even a transient addition of the anti-T790M drug at the beginning of treatment would prevent acquiring the resistance by the entire tumor and dramatically extend progression-free survival (PFS) and overall survival (OS). Osimertinib should be administered without waiting for the tumor to develop resistance, as it is feasible to eliminate a T790M-positive cell, but impossible to eradicate all million cells once the mutation is widespread. The latter scenario fails because of tumor heterogeneity, bad luck, and a mere probability. If the probability to kill one cell out of one is 0.99, then the probability to kill 2 cells out of 2 is 0.99 × 0.99 = 0.98, and the probability of killing all million cells is practically zero. Read more.