Advances in Treatment of Locally Advanced or Metastatic Non–Small Cell Lung Cancer

The treatment of metastatic non–small cell lung cancer (NSCLC) is constantly evolving. Although the advent of immunotherapy has played an important role in the treatment of patients with NSCLC, the identification of driver mutations and the subsequent specific treatment of these targets often lead to durable responses while maintaining quality of life. This review delves into targeted therapies available for epidermal growth factor receptor, anaplastic lymphoma kinase, ROS1, neurotrophic tropomyosin receptor kinase, and BRAF- mutated NSCLC patients, as well as other mutations with promising novel drugs under clinical investigation.

Key points

  • Most epidermal growth factor receptor (EGFR)-mutated NSCLC (exon 19 del and L858R) should be initially treated with osimertinib.

  • Anaplastic lymphoma kinase (ALK) fusion-positive NSCLC should be initially treated with alectinib, brigatinib, or ceritinib; however, tolerability issues limit the use of ceritinib.

  • ROS1 fusion-positive NSCLC should be initially treated with entrectinib over crizotinib given central nervous system activity.

  • BRAF V600E-mutated NSCLC should be initially treated with dabrafenib plus trametinib.

  • Neurotrophic tropomyosin receptor kinase fusion-positive NSCLC should be initially treated with entrectinib or larotrectinib.

  • Patients with HER2 or EGFR exon 20 insertions, RET fusions, NRG1 fusions, MET amplification and exon 14 skipping mutations, and KRAS G12C mutations should be initially treated with standard-of-care chemoimmunotherapy; however, there are targeted therapies under investigation showing promise.


The World Health Organization identifies lung cancer as the leading cause of cancer-related mortality and one of the highest-incidence cancer types. Within the 85% of all lung cancer that comprise non–small cell lung cancer (NSCLC), there is significant variability in the incidence of oncogenic driver mutations. For instance, within lung adenocarcinoma, 40% to 45% of tumors have identifiable potentially actionable mutations on top of the 25% of tumors with KRAS mutations, with higher numbers in strictly Asian populations. , Alternatively, patients with squamous cell carcinoma of the lung have less than a quarter of these mutations. The goal of molecular testing in NSCLC is to correctly identify and target oncogenic drivers. This review focuses on the advancements in treatment of oncogenically addicted locally advanced or metastatic NSCLC ( Fig. 1 , Tables 1 and 2 ).

Fig. 1

For patients with locally recurrent or metastatic disease requiring systemic therapy, this diagram indicates mutations with FDA-approved treatments available ( blue ) as well as mutations with associated agents under investigation ( red ). Agents are listed in alphabetical order. a Atypical EGFR mutations include S768I, L861Q, and G719.

Table 1

Common statistical acronyms used in clinical trials

Acronym Extended Form Meaning
CR Complete response Disappearance of all lesions
PR Partial response >30% decrease in the size of target lesions by RECIST 1.1
SD Stable disease Range of 30% decrease to 20% increase in the size of target lesions by RECIST 1.1
PD Progressive disease >20% increase in the size of target lesions by RECIST 1.1 and/or appearance of new lesions
PFS Progression-free survival Amount of time from randomization until a patient develops progressive disease or dies
OS Overall survival Amount of time from randomization until a patient dies from any cause
ORR Overall response rate The percentage of patients with CR + PR
DCR Disease control rate The percentage of patients with CR + PR + SD

RECIST (response evaluation criteria in solid tumors) v1.1 refers to a system of calculating responses to treatment based on the radiographic changes of measurements of target lesions in organs and lymph nodes.

Table 2

Names of common genes and proteins targeted in the treatment of non–small cell lung cancer

Gene/Protein Meaning
EGFR Epidermal growth factor receptor, both a gene and its receptor tyrosine kinase, also known as ERBB1
ALK Anaplastic lymphoma kinase, both a gene and its receptor tyrosine kinase
ROS1 Both a gene and its receptor tyrosine kinase
NTRK Neurotrophic tropomyosin receptor kinase (NTRK1, NTRK2, NTRK3) genes encode tropomyosin receptor kinases (TRKA, TRKB, TRKC)
BRAF A gene that encodes the B-Raf protein, with V600E the most effectively targeted oncogenic mutation
RET Both a gene and its receptor tyrosine kinase, it technically is an abbreviation for “rearranged during transfection”
MET Both a gene and its receptor tyrosine kinase, also known as c-Met
HER2 Both a gene and its receptor tyrosine kinase, also known as ERBB2
NRG1 Neuregulin 1, both a gene and a cell adhesion molecule that interacts with the ERBB receptor tyrosine kinases
KRAS A gene that encodes the K-Ras protein as part of the RAS/MAPK pathway
ERBB Both a gene and a family of 4 receptor tyrosine kinases, which include EGFR and HER2
MEK Proteins (MEK1 and MEK2) that are part of the MAPK pathway

Epidermal growth factor receptor

In a metaanalysis of 115,815 patients with NSCLC, 30,466 patients (32.3%) were found to have epidermal growth factor receptor (EGFR) mutations, the vast majority with exon 19 deletions or the exon 21 mutation L858R. Among Asian NSCLC patients, EGFR was detected in 38.8% compared with approximately 17% in Caucasians and African Americans with increased incidence also found in women, nonsmokers, and patients with adenocarcinoma histology.

A metaanalysis evaluating first-line first-generation EGFR tyrosine kinase inhibitor (TKI) therapy, erlotinib or gefitinib, compared to platinum doublet chemotherapy in EGFR-mutated NSCLC showed a median progression-free survival (PFS) and median overall survival (OS) of 11.0 months and 25.8 months versus 5.6 months and 26.0 months, respectively, with 73.8% of chemotherapy patients crossing over and receiving EGFR TKI therapy upon progression. Although no OS difference was seen, likely because of crossover, it was noted that median OS increased from 13.6 to 27.2 months in Japan after gefitinib was approved.

The advent of second-generation EGFR TKIs, afatinib and dacomitinib, with irreversible inhibition of EGFR as well as off-target effects on other ERBB receptors, fueled hope of improved efficacy. In a phase 2b LUX-Lung 7 trial comparing afatinib to gefitinib in treatment-naïve EGFR-mutated advanced NSCLC patients, overall response rate (ORR), median PFS, and median OS were 72.5%, 11.0 months, and 27.9 months with afatinib compared with 56.0%, 10.9 months, and 24.5 months with clinical significance reached only for ORR and median PFS. In the phase 3 ARCHER 1050 trial comparing dacomitinib to gefitinib in untreated EGFR-mutated NSCLC patients, median PFS and OS were 14.7 months and 34.1 months versus 9.2 months and 26.8 months, respectively, both significant, with nonsignificant differences in ORR.

Second-generation EGFR TKIs were soon eclipsed by osimertinib, a third-generation EGFR TKI with efficacy against the exon 20 T790M mutation, the gatekeeper, and preeminent secondary EGFR resistance mutation. The phase 3 AURA-3 trial evaluated osimertinib versus platinum-based pemetrexed doublet in 419 patients with EGFR T790M mutations after first-line EGFR TKI therapy. ORR was 71% with osimertinib compared with 31% with chemotherapy. Median PFS was improved with osimertinib at 10.1 months versus 4.4 months. In the phase 3 FLAURA study of 556 untreated EGFR-mutated NSCLC patients randomized to osimertinib versus gefitinib or erlotinib, ORR, median PFS, and OS were significantly improved at 80%, 18.9 months, and 38.6 months compared with 76%, 10.2 months, and 31.8 months, respectively, even with 43% crossover in the first-generation arm. ,

Although these data solidified osimertinib as the standard first-line therapy for EGFR exon 19 and 21 mutations, many studies disregarded exon 20 insertions and other rare point mutations. Although A763_Y764FQEA accounts for 10% to 20% of EGFR exon 20 mutations and appears in vitro to be responsive to gefitinib and erlotinib, the other 80% to 90% are resistant to first-generation EGFR TKI therapy, and both these and rare point mutations have not been studied with osimertinib.

In a post hoc analysis of afatinib versus chemotherapy in LUX-Lung 2, 3, and 6, rare mutations were evaluated. Atypical point mutations in exon 18 to 21 showed an ORR of 71.1%, median PFS of 10.7 months, and median OS of 19.4 months. Exon 20 insertions showed an ORR of 7.1%, median PFS of 2.7 months, and median OS of 9.2 months.

Although there is promise with standard EGFR TKI therapy for atypical EGFR point mutations, EGFR exon 20 insertions exhibit exceedingly poor responses to standard TKIs. AUY922, a heat shock protein 90 inhibitor, was evaluated in a small study of 10 EGFR exon 20 insertion NSCLC patients with a median PFS of 6.1 months. Poziotinib, a smaller TKI with binding potential within the altered exon 20 insertion kinase pocket, demonstrated an ORR of 58%, disease control rate (DCR) of 90%, and median PFS of 5.6 months. TAK-788, another TKI with activity against EGFR and HER2 in NSCLC, is being investigated in a phase 1/2 study with a preliminary ORR of 21% and DCR of 64% in 14 patients.

With regards to resistance mechanisms to osimertinib therapy, the most commonly detected alterations include MET amplification, HER2 amplification, secondary resistance mutations (namely C797S), and secondary oncogenic drivers. In the TATTON phase 1b expansion study of patients with EGFR mutant NSCLC after EGFR TKI therapy who developed MET amplification, the combination of osimertinib and savolitinib, a MET inhibitor, produced an ORR of 20% in 25 patients. With regards to HER2 amplification as a resistance mechanism, preclinical studies have suggested the addition of T-DM1 to osimertinib can significantly suppress tumor growth of resistant cells. Data support the treatment of T790M in trans with C797S with the combination of osimertinib and first-generation EGFR TKI. Finally, multiple case reports have reported the development of secondary oncogenic drivers after EGFR TKI therapy, with new anaplastic lymphoma kinase (ALK) and RET fusions detected responding to combinations of osimertinib and alectinib as well as afatinib and cabozantinib, respectively. These resistance mechanisms with responses to additional targeted therapy highlight the benefit of obtaining postprogression biopsies.

Anaplastic lymphoma kinase

Although ALK was originally identified in anaplastic large cell lymphoma, the fusion of EML4 and ALK via the inversion of chromosome 2p was the first identification of ALK in NSCLC. Numerous ALK fusions have been identified, which lead to the constitutive activation of the ALK kinase domain and subsequent downstream pathways, including PI3K-AKT, JAK-STAT3, and ERK/MAPK. Incidence of ALK fusions in NSCLC varies, ranging from 3% to 13% of cases, with The Lung Cancer Mutation Consortium identifying 8% of 733 adenocarcinomas tested with ALK rearrangements. ,

The phase 3 PROFILE 1014 examined the treatment of 343 therapy-naïve ALK rearranged NSCLC patients comparing crizotinib to pemetrexed-based platinum doublet chemotherapy. The ORR and PFS were 74% versus 45% and 10.9 months versus 7.0 months, respectively. The 4-year OS was 56.6% versus 49.1% seeming to favor crizotinib, although was nonsignificant given 84.2% of patients in the chemotherapy arm subsequently received crizotinib.

The second-generation ALK TKIs, ceritinib, alectinib, and brigatinib, were investigated to improve upon crizotinib in the treatment of ALK rearranged NSCLC. Ceritinib was investigated in ASCEND-5 in ALK rearranged NSCLC patients who had previously received chemotherapy and crizotinib compared with standard chemotherapy (pemetrexed or docetaxel). Even with prior ALK TKI therapy, ORR, DCR, and median PFS in the ceritinib group were 45%, 88%, and 5.4 months compared with 8%, 42%, and 1.6 months in the chemotherapy group. In ASCEND-4, treatment-naïve patients received ceritinib versus pemetrexed-based platinum doublet therapy; the ORR was 72.5% and median PFS was 16.6 months compared with 26.7% and 11.1 months, respectively.

Two phase 3 trials compared alectinib to crizotinib. ALEX randomized 303 treatment-naïve ALK rearranged patients to alectinib 600 mg orally twice a day versus crizotinib and showed a median PFS of 34.8 months versus 10.9 months and an ORR of 82.9% versus 75.5%, respectively. Twelve-month OS was similar at 84.3% and 82.5%. J-ALEX randomized 207 Japanese patients who previously received up to 1 prior treatment with chemotherapy to alectinib 300 mg orally twice a day versus crizotinib and showed similar initial results with an ORR of 85% versus 70% and median PFS not reached versus 10.2 months, respectively.

In the phase 2 ALTA study of 222 ALK rearranged NSCLC patients who previously received crizotinib or chemotherapy, brigatinib showed an ORR of 53%, DCR of 92%, median PFS of 15.6 months, and 12-month OS of 80%. The subsequent ALTA-1L phase 3 study of 275 patients without previous ALK-based therapy who had received up to 1 chemotherapy evaluated brigatinib versus crizotinib. Preliminary data show an ORR, 12-month PFS, and 12-month survival without intracranial disease progression in those with baseline brain metastases of 71% versus 60%, 67% versus 43%, and 67% versus 21% favoring brigatinib.

The third-generation ALK TKI lorlatinib has shown significant benefit in ALK rearranged patients regardless of treatment history. In the phase 2 expansion study, patients were assigned to different ALK-positive arms ranging from treatment-naïve to previous treatment with up to 3 different ALK TKIs. In the treatment-naïve cohort, the ORR was 90% and intracranial response was 66.7%. In patients previously only treated with crizotinib, ORR was 69.5% and intracranial response was 87%. In patients treated with between 1 and 3 previous TKIs, the ORR was 47% with an intracranial response of 63% indicating benefit in heavily pretreated patients.

As a mechanism of resistance, secondary ALK point mutations arise in order to interfere with ALK-directed TKI binding under selective pressure from TKI therapy, ultimately allowing continued constitutive binding. The gatekeeper mutation L1196M was identified in conjunction with C1156Y, after a patient progressed on therapy with crizotinib. Another set of secondary mutations, G1202R, S1206Y, and 1151Tins, were detected after a patient progressed on crizotinib. Although S1206Y is sensitive to treatment with any second-generation ALK TKI, and 1151Tins is sensitive to alectinib and brigatinib, notably, G1202R produces resistance to crizotinib, ceritinib, alectinib, and brigatinib that can only be overcome by lorlatinib.


The ROS1 gene encodes a receptor tyrosine kinase related to ALK, LTK, and the insulin receptor family. ROS1 fusions are estimated to occur in 1% to 2% of NSCLC and were initially identified in glioblastoma via the FIG-ROS1 fusion generated by the intrachromosomal homozygous deletion on 6q21 with subsequent downstream phosphorylation of SHP-2 and the PI3K-AKT, JAK-STAT3, and ERK/MAPK pathways. ,

PROFILE 1001, an expansion cohort of the evaluation of crizotinib as an MET and ALK inhibitor, was the first prospective evaluation of crizotinib in ROS1 rearranged NSCLC. Fifty largely pretreated with chemotherapy TKI-naïve ROS NSCLC patients received crizotinib with an ORR of 72%, DCR of 90%, median PFS of 19.3 months, and median OS of 51.4 months. In a subsequent East Asian phase 2 study in 127 pretreated TKI-naïve ROS1 NSCLC patients, crizotinib demonstrated an ORR of 71.7%, DCR of 88.2%, median PFS of 15.9 months, and 12-month OS of 83.1%. When stratified by baseline brain metastases, the median PFS was 10.2 months with brain metastases and 18.8 months without.

Several phase 2 studies have evaluated ceritinib, lorlatinib, and entrectinib (a multitargeted ALK, ROS1, and NTRK1/2/3 TKI) in pretreated ROS1-positive NSCLC. Although ceritinib resulted in mixed results, in the subset of the 30 TKI-naïve patients, the ORR was 67%, DCR was 87%, and median PFS was 19.3 months. In a study that enrolled 59 patients with ROS1 fusions, 38 previously receiving ROS1 TKI therapy, lorlatinib resulted in an ORR of 29.4%. For TKI-naïve patients, the ORR was 76.9%. An abstract combining the phase 1 studies and the ongoing phase 2 study STARTRK2 of entrectinib showed in 53 ROS1 NSCLC TKI-naïve patients an ORR of 77%, median PFS of 26 months in patients without baseline central nervous system (CNS) disease, and 14 months with CNS disease.

Two newer treatments have been evaluated with phase 1 studies for the treatment of ROS1 NSCLC: repotrectinib (a next-generation ALK, ROS1, and NTRK1/2/3 TKI) and DS-6051b (a high-affinity TKI for ROS1 and NTRK1/2/3). Repotrectinib was investigated in TRIDENT-1 with 31 ROS1 rearranged tumors, 29 NSCLC, showing an ORR of 70% among ROS1 TKI-naïve patients and 11% among TKI refractory patients. In the US study of DS-6051b, the ORR was 33.3% and DCR was 66.7% in 9 patients, 7 of whom were previously treated with crizotinib. In the Japanese study of DS-6051b, the ORR was 58.3% for the overall cohort, 66.7% in the TKI-naïve cohort, and 33.3% in the TKI pretreated cohort, with a DCR of 100%.

To date, there are 3 secondary ROS1 mutations, G2032R, D2033N, S1986Y/F, that have been identified in patients who generate resistance to ROS1 TKIs with numerous other mutations, including the gatekeeper L2026M, only identified in vitro. G2032R was identified in a CD74-ROS1 NSCLC patient with evidence suggesting resistance to crizotinib and ceritinib but sensitivity to cabozantinib, lorlatinib, and repotrectinib. , D2033N was also identified in a CD74-ROS1 NSCLC patient with evidence suggesting resistance to crizotinib, ceritinib, brigatinib, lorlatinib, but sensitivity to cabozantinib and repotrectinib. , S1986Y/F were identified in an EZR-ROS1 NSCLC patient with evidence suggesting resistance to crizotinib and ceritinib, but sensitivity to lorlatinib.

Neurotrophic tropomyosin receptor kinase

NTRKs play a physiologic role in CNS development, in which they respond to Nerve Growth Factor, Brain-Derived Neurotrophic Factor, and Neurotrophin-3. , Transmembrane receptors TRKA, TRKB, and TRKC are encoded by the genes NTRK 1, 2, and 3, respectively. Important downstream effectors of NTRK include the RAS/RAF/MAPK/ERK, PI3K/AKT/mTOR, and PLCy/PKC pathways and are implicated in cellular transformation, proliferation, migration, and invasiveness. In NSCLC, NTRK1 fusions are identified in approximately 1% of patients.

Efficacy of entrectinib was investigated in several trials. A total of 119 patients were treated with entrectinib, of whom 71 had NSCLC, and pooled outcomes from STARTRK-1, STARTRK-2, and ALKA-372-001 looked at efficacy outcomes in 54 adult patients with advanced or metastatic tumors demonstrating NTRK fusions across 10 tumor types. After 15.5 months of follow-up, ORR was 57.4% with responses in all included tumor types. Median duration of response was 10.4 months, median PFS was 11.2 months, and median OS was 20.9 months. Larotrectinib was also recently approved in a tumor-agnostic manner for NTRK-related cancers after showing an ORR of 75% with 13% obtaining stable disease. Of note, the 55 patients in this study represented 17 different malignancies, with only 4 patients having lung cancer.

Ongoing clinical trials in NTRK include repotrectinib (phase 1, 2, NTRK 1–3, NCT03093116 ), and cabozantinib in NTRK mutation and RET fusion (phase 2, NCT01639508 ). Other compounds with NTRK activity include TSR-011, PLX-7486, XL-184, MGDC516, DS6051b, F17752, and DCC-2701.


BRAF is a serine/threonine kinase regulating cell growth in the RAS-RAF pathway. BRAF mutations, particularly V600E, are found in a variety of human cancers, most commonly malignant melanoma. In NSCLC, the prevalence of BRAF V600E is 1% to 3%. , The BRAF V600E mutation induces activation of downstream MEK-ERK signaling resulting in cell proliferation.

BRAF V600E inhibitors dabrafenib and vemurafenib have been studied as single agents in NSCLC. In 1 study, 84 patients with metastatic BRAF V600E-mutated NSCLC without brain metastases either with or without previous treatment were started on dabrafenib. Among 78 pretreated patients, the ORR was 33%. Median OS was 12.7 months. Among 6 previously untreated patients, 4 patients had partial responses with PFS ranging from 4.5 to 16.6 months.

In 1 basket study, patients with nonmelanoma cancers with BRAF V600E received vemurafenib. Among 20 patients with NSCLC, the ORR was 42% and median PFS was 7.3 months.

MAPK pathway inhibition with combination of dabrafenib and MEK inhibitor trametinib has been studied in patients with pretreated and previously untreated stage IV NSCLC with BRAF V600E mutation. Among 57 pretreated patients, the ORR was 63.2% with a median PFS of 9.7 months. Among 36 patients who had previously untreated BRAF V600E-mutated NSCLC, the ORR was 64%.


RET is a protooncogene located on chromosome 10q11.2 that encodes a receptor tyrosine kinase involved in neural crest and enteric ganglia development through the impact of its downstream effectors RAS/MAPK/ERK, PI3K/AKT, and JAK/STAT on cellular differentiation, migration, and proliferation.

In 2012, surgical resections from 936 patients with NSCLC were evaluated by polymerase chain reaction for RET fusion, validated by immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH). Thirteen tumors (1.4%) had exclusive RET fusions. A separate 2013 analysis of 1874 NSCLC tumors at a large Japanese center confirmed these findings, showing an incidence of 1.2% in NSCLC.

TKIs active against RET fusions are often multikinase inhibitors already in clinical use. A global registry study from 2015 to 2016 identified 165 patients with RET-rearranged NSCLC from the United States, Europe, and Asia, of whom 53 previously TKI-naïve patients received at least 1 line of multikinase inhibitor during the registry period (cabozantinib, vandetanib, sunitinib, sorafenib, alectinib, lenvatinib, nintedanib, ponatinib, and regorafenib). ORRs were cabozantinib 37%, vandetanib 18%, and sunitinib 22%. Median PFS was 2.3 months. Median duration of therapy was 1.8 months.

A 26-patient, phase 2 single-arm trial of cabozantinib in RET-rearranged NSCLC from 2012 to 2016 attained an ORR of 28%. An 18-patient phase 2 trial of vandetanib in RET-rearranged NSCLC from 2013 to 2015 attained an ORR of 18%. A separate, 19-patient phase 2 trial of vandetanib attained an ORR of 53%. Small case series with alectinib and sorafenib , described partial responses to these drugs.

Ongoing phase 1 and 2 clinical trials of RET-specific agents include selpercatinib with preliminary ORR of 68% and pralsetinib with preliminary ORR of 50%. Recently published phase 1 data on another RET-multikinase inhibitor, RXDX-105, has shown an ORR of 19%, with 0/20 objective responses in KIF5B-RET, whereas other fusion partners responded in 6/9 cases.

Nakaoku and colleagues reported secondary resistance to vandetanib in a patient with metastatic lung adenocarcinoma and CCDC6-RET fusion, resulting from an acquired serine-to-phenylalanine substitution at codon 904 of the RET kinase domain and supported this finding in vitro. Nelson-Taylor and colleagues reported resistance to ponatinib in RET-rearranged lung adenocarcinoma cell lines via NRAS and EGFR/ALK alterations that restored RAS/MAPK signaling.

Met amplification and met exon 14 skipping mutations

MET, a receptor tyrosine kinase, together with its ligand hepatocyte growth factor/scatter factor mediates survival and long-distance migration of epithelial and myogenic precursor cells during embryogenesis. In cancer, it is implicated as a mechanism of invasion and metastasis, thus is an attractive therapeutic target. Gain-of-function mutation of MET can occur by gene amplification or by MET exon 14 skipping mutations (METex14), which impairs degradation of MET receptors. The prevalence of MET amplification (note there are different scoring systems defining MET amplification) is reported to be 4% to 5%, whereas that of METex14 is 2% to 3%. ,

There are currently no approved therapies targeting NSCLC with MET amplification or mutation. Various MET inhibitors (crizotinib, cabozantinib, tepotinib, telisotuzumab vedotin, capmatinib, and savolitinib) are under active investigation, and preliminary studies show promising results.

In the phase 2 AcSé trial, patients with advanced NSCLC with c-MET ≥ 6 copies, c-MET-mutated, or ROS1 translocated tumors were treated with crizotinib. Twenty-five patients were included in the c-MET ≥ 6 copies cohort, 28 in the c-MET-mutation cohort (25 of which had exon 14 mutation), and 37 in the ROS1 translocation cohort. ORRs after 2 cycles were 16%, 10.7%, and 47.2%, respectively, the median PFS was 3.2, 2.4, and 5.5 months, respectively, and the median OS was 7.7, 8.1, and 17.2 months, respectively.

In a retrospective study of 148 patients with metastatic METex14 NSCLC, 27 patients received at least 1 MET inhibitor. The median OS among those patients was 24.6 months. MET inhibitor as any line of therapy was associated with prolongation of survival with a hazard ratio of 0.11.

Other ongoing phase 2 trials include treatment with cabozantinib ( NCT01639508 and NCT03911193 ), tepotinib ( NCT02864992 and NCT03940703 ), telisotuzumab vedotin ( NCT03539536 ), capmatinib ( NCT03693339 , NCT02414139 , NCT01610336 , and NCT02750215 ), crizotinib ( NCT04084717 ), and savolitinib ( NCT03778229 ).


The ERBB2 gene encodes HER2, which signals through the PI3K-AKT and MEK-ERK pathways. In-frame insertions in exon 20 lead to constitutive activation. Protein overexpression and gene amplification are present in up to 35% of NSCLC; however, mutations are less common, occurring in less than 4% of NSCLC. ,

Studies evaluating HER2 overexpression by IHC or amplification by FISH in NSCLC as a target for adding trastuzumab to cisplatin and gemcitabine , did not show evidence of clinical benefit. Attention was turned to somatic mutations in HER2 kinase, which were estimated to be present in 1.5% of NSCLC, although higher in women and nonsmokers. A high degree of similarity among these mutations was noted, with exon 20 insertions accounting for 96%, and most affecting codon 775.

A retrospective analysis of 3800 patients identified 65 patients (1.7%) with exon 20 HER2 insertions in lung adenocarcinoma. Twenty-two courses of HER2-directed therapy were given to 16 patients, with an ORR of 50% and a DCR of 82%. Responses were seen with trastuzumab (15 treatments) and afatinib (3 treatments). A phase 2 trial of dacomitinib in patients with exon 20 mutated or HER2-amplified NSCLC achieved partial responses in 3 of 26 patients with exon 20 insertions, and 0 of 4 patients with HER2 amplification.

A retrospective study was performed of 27 patients with stage IV or recurrent HER2-mutated NSCLC who were treated with afatinib. Of 23 evaluable patients, 3 had partial response to afatinib, including 2 who had previously been treated with trastuzumab and pertuzumab. Another 13 patients had stable disease. The median duration of response to afatinib was 6 months.

In the EUHER2 study of 101 patients with lung adenocarcinoma and HER2 exon 20 insertions who had undergone a median of 3 lines of treatment, inclusive of both conventional chemotherapy and HER2-directed agents (trastuzumab, neratinib, afatinib, lapatinib, T-DM1), an ORR of 50.9% and PFS of 4.8 months were achieved in 58 patients who received trastuzumab or T-DM1. By comparison, ORR was 43.5% and PFS was 6 months for conventional chemotherapy in the first line, and 10% and 4.3 months in the second line. EGFR TKI resulted in an ORR of 7.6% and PFS of 2.99 months. HER2 TKI (neratinib, lapatinib, afatinib) showed an ORR of 7.4% and PFS of 3.4 months. A separate retrospective study evaluated afatinib in 16 patients with metastatic NSCLC, exon 20 insertion, and a median of 3 lines of therapy. The ORR was 19% and the DCR was 69%. A phase 1 trial of the combination of neratinib and temsirolimus in multiple tumor types noted partial responses in 2 of 5 NSCLC with exon 20 insertion.

Trastuzumab emtansine, a HER2-directed antibody-drug conjugate, has been shown to have limited efficacy in NSCLC. Contrarily, ado-trastuzumab emtansine was shown to result in a partial response rate of 44% and a median PFS of 5 months in a study of 18 patients with advanced lung adenocarcinoma harboring HER2 mutation. Ongoing phase 1 and 2 trials in HER2-directed therapy include TAK-788 (phase 1, 2), poziotinib, neratinib, and trastuzumab deruxtecan (DS-8201, Daiichi phase 2 NCT03505710 ).

Neuregulin 1

Mutations in the ERBB family are important drivers in NSCLC. In addition to ERBB1 (EGFR) and ERBB2 (HER2), ERBB3 and ERBB4 play an indirect role in the oncogenesis of invasive mucinous adenocarcinomas (IMAs) through neuregulin-1 (NRG1) fusions. Upon binding a ligand, ERBB1-4 form homodimers or heterodimers to induce phosphorylation of the kinase domain and ultimately downstream signaling through PI3K-AKT and MAPK pathways.

The fusion of NRG1, located on chromosome 8, with various partners creates an extracellular NRG1 ligand bound to the partner intracellular domain, which tethers the NRG1 ligand and leads to repeated binding to ERBB3 and downstream activation. ,

In a retrospective analysis of 21,858 solid tumor specimens, the overall incidence of NRG1 fusions was 0.2% across all tumor types with the vast majority of NRG1 fusions in NSCLC (25 of 41 samples). Within NSCLC, 32% of NRG1 fusions were found within tumors with mucinous features, and they largely were adenocarcinoma. IMAs are estimated to constitute 2% to 10% of all lung adenocarcinomas and have historically presented as aggressive multifocal malignancies with poor responses to chemotherapy. , The overall incidence of NRG1 fusions in IMAs is estimated between 7% and 27%. , ,

Although there are currently no Food and Drug Administration (FDA) -approved drugs specifically for NRG1 fusion-positive NSCLC, afatinib has shown variable efficacy. Drilon and colleagues reported the use of afatinib in 3 therapy-naïve NRG1 fusion-positive NSCLC patients with no meaningful response. In addition, both Drilon and colleagues and Kim and colleagues attempted the use of afatinib after treating patients with ERBB3-directed therapy with no response. However, responses were reported with afatinib use in 4 separate patients ranging from 6.5 to 12 months. Alternatively, 1 patient treated with the combination of lumretuzumab, an ERBB3 monoclonal antibody, and erlotinib exhibited stable disease for months. Finally, 1 patient treated with GSK2849330, an ERBB3 monoclonal antibody, produced a partial response to therapy for 19 months.


KRAS mutations are one of the most common mutations found in NSCLC, reported in 20% to 30% of cases. , In NSCLC, multiple mutations of the KRAS gene have been described, in codons 12, 13, and 61. Rather than representing a homogenous disease, KRAS-mutated NSCLC appears to be a diverse set of diseases with other associated mutations. Unfortunately, inhibition of KRAS alone or in combination with MEK inhibitors has yielded disappointing results.

Recently presented phase 1 data of AMG 510, a KRAS G12C inhibitor, has shown promising results. Patients with locally advanced or metastatic NSCLC with a KRAS G12C mutation were enrolled. Among 13 evaluable patients who have reached target dosing, 7 patients achieved partial response and the other 6 patients exhibited stable disease. Another phase 1/2 trial of KRAS G12C inhibitor, MRTX849, in advanced solid tumors was recently presented. Among 6 patients with evaluable responses, 3 had partial responses and 3 had stable disease. If effective, these will be the first successful targeted therapies against KRAS in the 20 years the KRAS mutation has been known.


Since the early 2000s, the identification of targetable mutations in NSCLC has allowed for the treatment of patients with advanced or metastatic disease with medications, namely TKIs, designed to block the oncogenic driver and delay chemotherapy. FDA-approved medications now exist to treat NSCLC with mutations in EGFR, ALK, ROS1, NTRK, and BRAF with newer drugs being developed to treat resistant disease. In addition, targeted therapies are being developed and tested for numerous targets, including RET, MET, KRAS, NRG1, and exon 20 insertions of HER2 and EGFR. The durable responses many patients experience with these agents make the case for sending broad mutational tissue testing.


L. Bazhenova has received advisory board honoraria from the following companies in the past 12 months: Blueprint, Takeda, Astra Zeneca, Genentech, BeyondSpring Pharmaceuticals, G1 Therapeutics, Boehringer Ingelheim. L. Bazhenova has received research funding from BeyondSpring Pharmaceuticals .


Only gold members can continue reading. Log In or Register to continue

Aug 16, 2020 | Posted by in GENERAL | Comments Off on Advances in Treatment of Locally Advanced or Metastatic Non–Small Cell Lung Cancer

Full access? Get Clinical Tree

Get Clinical Tree app for offline access