Chemical Reviews pubs.acs.
org/CR Review
EGFR.228 Icotinib has also been approved by the China Food examples which were not able to successfully target the
and Drug Administration (CFDA) for the treatment of T790M mutation in clinical trials. An early publication with
NSCLC and is also generally referred to as a first-generation small molecule inhibitors to support this hypothesis was
molecule.229 The response rates of patients with EGFR disclosed in 2009 from the DFCI group,234 exemplified by
mutations to these treatments is encouraging at around 70%, WZ-4002 in Figure 37. Key to the hypothesis of the profile is
but resistance typically develops within 9−12 months. Most
commonly, the disease progression is driven by the emergence
of a secondary acquired mutation, T790M (i.e., threonine
gatekeeper mutated to methionine), in combination with an
activating sensitizing mutation.230 The mutation of the
gatekeeper residue is well precedented in a number of the
case studies covered in this review including ALK, BCR/ABL,
and cKIT. This inhibitor-resistant form of mutant EGFR is
commonly referred to as the “double mutant” (DM) form but
can consist of either activating mutant.
The specific mechanism by which these activating mutations
of EGFR exert their effect is still a focus of research efforts, but
in vitro studies on the L858R mutation have shown a decreased
affinity toward ATP (Figure 36).80 This also has the result of
increasing the potency of first-generation inhibitors such as
gefitinib and erlotinib. It has been speculated that the
activating mutations bias the equilibrium toward the active
“c-helix in” conformation of the kinase to which the inhibitors
bind. Further reports suggest additional mechanisms, including
the impact on the release of ADP. The acquired mutation,
T790M, which results in resistance to first-generation
inhibitors, has been shown to increase the affinity of the
kinase for ATP, substantially reducing the cellular activity of
these ATP-competitive inhibitors. The mechanism of this
mutation appears distinct to other kinases where the impact
occurs primarily due to steric impact on inhibitor binding.
However, it is also likely that there is a steric component to the
resistance mechanism against first-generation EGFR inhibitors,
which preferentially bind to the threonine gatekeeper.
Second-generation EGFR inhibitors were subsequently
developed based on the same anilinoquinazoline template;
these however have an irreversible covalent mechanism of
action. They contain an acrylamide group which covalently Figure 37. Early reported mutant selective “third-generation” T790M
bonds to C797.231 Two such examples evaluated in clinical inhibitors.
trials include afatinib232 and dacomitinib,233 which were
approved by the FDA for the treatment of EGFR mutant
NSCLC patients in 2013 and 2018 retrospectively. Although the exploitation of T790M to leverage selectivity against the
not initially designed for this purpose, the formation of the wild-type form. Parallel to this work, AstraZeneca discovered
covalent bond results in the ability of the compounds to T790M mutant selective inhibitors via focused biochemical
overcome the increased ATP affinity of the T790M mutant profiling to identify a reversible molecule (Figure 37) and
resulting in more potent activity in cellular models. This converted this into a covalent inhibitor using structure-based
cysteine, however, is present in all of these clinically relevant drug design (SBDD).235 Interestingly, not only were such
mutants of EGFR, and consequently second-generation compounds active against the T790M double mutant forms,
inhibitors have increased activity against wild-type (WT) they also maintained potent activity against the activating
EGFR. The inhibition of EGFR WT does not appear to mutations alone. Other research groups such as Avila/Celgene
contribute to clinical efficacy, but it is responsible for the also reported compounds targeting T790M exemplified by
commonly observed side effects, such as skin rash and diarrhea. rociletinib (formerly CO-1686).236
These side effects are often dose limiting in patients, and The third-generation mutant selective inhibitor osimertinib
therefore the wild-type activity of these inhibitors limits the (formerly AZD9291) entered clinical testing in 2013 and, due
efficacy which can be achieved against the T790M resistant to robust clinical efficacy, received breakthrough designation in
mutation in a clinical setting. 2014 and early FDA approval in 2015 with a PFS of 10.1
Alternative approaches to target the T790M mutation were versus 4.4 months against the standard of care. The discovery
therefore required; such inhibitors are now commonly referred story of osimertinib has been published and includes the
to as “third-generation” inhibitors. The concept of this profile disclosure of preclinical efficacy against the single and double
was to identify inhibitors which were “mutant-selective”, i.e., mutant forms along with a margin to wild-type EGFR.237,238
which were able to target the relevant mutant forms but with a The binding mode of osimertinib was modeled using
margin over wild-type EGFR. Such a profile would be distinct computational approaches, and from this the T790M mutant
to the previously reported first- and second-generation selective profile can be rationalized (Figure 38). Osimertinib
AC https://dx.doi.org/10.1021/acs.chemrev.0c00383
Chem. Rev. XXXX, XXX, XXX−XXX
