Epidermal growth factor receptor –mutated non–small cell lung cancer: a clinical approach

CLINICAL PEARLS

All patients with advanced, nonsquamous, non−small cell lung cancer (NSCLC) should undergo epidermal growth factor receptor (EGFR) mutation testing as a part of a broader molecular panel (e.g., targeted next-generation sequencing panel).
EGFR tyrosine kinase inhibitors (TKIs) are standard first-line therapy for advanced EGFR -mutated NSCLC, with >2 decades of data showing an improvement in progression-free survival when compared with platinum-based chemotherapy.
Osimertinib, a third-generation EGFR TKI, when available, is the frontline standard of care for advanced EGFR -mutated NSCLC. This is based on the FLAURA trial, which demonstrated an improvement in progression-free survival, overall survival, and improved central nervous system activity when compared with first-generation EGFR TKIs (i.e., erlotinib or gefitinib).
Liquid biopsies, which detect and measure circulating tumor DNA, provide noninvasive diagnostics for EGFR mutation and other molecular testing. However, liquid biopsies may be limited by the sensitivity of the assay. If the test is negative for the presence of a mutation, a confirmatory tissue biopsy should be obtained.
Avoid use of immunotherapy, either monotherapy or in combination with chemotherapy, in EGFR -mutated NSCLC. There is lack of data for use of immunotherapy in combination with chemotherapy and also data highlighting the poor clinical activity with immunotherapy monotherapy. This is critical as PD-L1 immunohistochemistry testing results may be available before molecular testing results are available.

Introduction and epidemiology

The discovery of epidermal growth factor receptor (EGFR) mutations has revolutionized the management of advanced lung cancer. Over the past 2 decades, the understanding of EGFR mutations, along with other molecular aberrations (e.g., ALK, ROS1, BRAF ), has resulted in a myriad of targeted therapy options. EGFR mutations are present in 10%–15% of non–small cell lung cancer (NSCLC) in Western populations and as high as 50%–60% in Asian populations. While classically associated with a never-smoking patient population, the incidence of EGFR mutations may be as high as 37% in regular smokers in Asian cohorts. This chapter uses a clinical lens to focus on the current data that support the diagnosis and treatment paradigm of EGFR -mutated NSCLC. Data from clinical trials are presented in the following order: hazard ratio (HR), with 95% confidence interval (CI) and P value for the comparison of two groups. We emphasize the importance of accurate molecular testing methods (i.e., blood and tissue), repeating molecular testing at the timing of resistance and considerations when sequencing available treatment options.

Molecular pathways and mutations

The EGFR signaling pathway, crucial for regulating tumorigenesis and cell survival, is overexpressed in the development and progression of NSCLC. EGFR exists as a monomer on the cell surface and thus must dimerize to activate tyrosine kinase activity. Activating somatic mutations occur in the region of the EGFR gene that encodes the tyrosine kinase domain (exons 18–24), resulting in constitutive activation of EGFR and leading to enhanced cell proliferation. The two most common EGFR activating mutations are exon 19 in-frame deletions and exon 21 L858R substitution mutations, which comprise ∼80%–90% of all EGFR mutations found in NSCLC. These EGFR mutations are also considered “sensitizing” mutations, which means that they are susceptible to currently available EGFR tyrosine kinase inhibitors (TKIs). EGFR TKIs are molecules that can block binding of adenosine-5-triphosphate to the constitutively activated EGFR tyrosine kinase catalytic domain, interrupting the aberrant signaling pathway and therefore preventing cell cycle progression and proliferation. ^,

Across the most common EGFR sensitizing mutations, the objective response rates (ORRs) are high (∼60%) with EGFR TKI therapy. Data from trials comparing EGFR TKI therapy to platinum-doublet chemotherapy suggest that patients with the EGFR exon 19 deletion may have better outcomes compared with those patients with EGFR exon 21 L858R substitution. In fact, in a preplanned combined analysis of two trials of 709 patients examining EGFR TKI afatinib (a second-generation EGFR TKI) compared with platinum-doublet chemotherapy, the median overall survival (OS) was not improved in using afatinib; however, in the subgroup analysis by mutation type, there was a statistically significant benefit in OS in only those patients with an EGFR exon 19 deletion. This is different from the first-generation EGFR TKIs, which in separation of the two common sensitizing mutations do not show an OS benefit.

In a meta-analysis of seven eligible trials examining first- and second-generation EGFR TKIs, the HR for progression-free survival (PFS) favoring EGFR TKI therapy over chemotherapy was stronger in patients with EGFR exon 19 deletions (HR, 0.24; 95% CI, 0.20−0.29) compared with those with EGFR exon 21 L858R substitutions (HR, 0.48; 95% CI, 0.39−0.58; P value for interaction <0.001). Interestingly, more recent data from the phase III FLAURA study, which examined third-generation EGFR TKI (i.e., osimertinib) compared with earlier generation TKIs (i.e., gefitinib or erlotinib), also showed a trend for improvement in both PFS and OS in those with EGFR exon 19 deletions, compared with those with L858R substitutions, (exon 19 deletion: PFS HR 0.43 (95% CI, 0.32–0.56) and OS HR 0.68 (95% CI, 0.51–0.90); L858R PFS HR 0.51 (95% CI, 0.36–0.71) and OS HR, 1.00 (95% CI, 0.71–1.40). This key trial is described in further detail as follows. However, this differential effect has not been consistent in all studies of inter-EGFR TKI comparisons. Emerging data suggest that even the subtype of EGFR exon 19 deletion (i.e., L747-A750>P) may matter for EGFR TKI sensitivity.

There are also uncommon sensitizing EGFR mutations (e.g., exon 18 G719X, L861Q; exon 20 S768I) accounting for ∼10%–12% of all EGFR mutations. There are less data to support a standard approach for treatment in patients with these less common mutations, although they may have worse outcomes with EGFR TKI therapy (discussed in further detail later). Of note, not all EGFR mutations are sensitive to currently available EGFR TKIs. For example, in-frame base pair insertions in exon 20 also result in the activation of EGFR and comprise ∼3%–5% of all EGFR -mutated NSCLC. However, most of these mutations are associated with primary EGFR TKI resistance, with only a ∼3%–8% response rate and a median PFS of ∼2 months. Table 10.1 shows a summary of the most used TKIs and their key properties, relevant for clinical use.

TABLE 10.1 ■

Properties of Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitors

Name	EGFR Binding Type	Receptor Targets	Half Life (hours)	Food Effect	CNS Penetration
Erlotinib (first generation)	Reversible	EGFR/HER1	36	Increased plasma concentration, take on empty stomach	0.03× CSF/Plasma
Gefitinib (first generation)	Reversible	EGFR/HER1	48	No change	0.01× CSF/Serum
Afatinib (second generation)	Irreversible	Pan-HER inhibitor (EGFR/HER1, HER2, HER4, indirect HER3 activity)	37	Decreased plasma concentration, take on empty stomach	0.02× CSF/Plasma
Dacomitinib (second generation)	Irreversible	Pan-HER (EGFR/HER1, HER2, HER4)	59–85	No change	Data unavailable
Osimertinib (third generation)	Irreversible	EGFR/HER1, HER2, HER3, HER4	48	No change	2× Brain/plasma

CSF, Cerebrospinal fluid; EGFR, epidermal growth factor receptor.

Molecular testing at diagnosis

Routine testing for EGFR mutations, as part of broader molecular panel, is recommended in all advanced or recurrent nonsquamous NSCLC. The international standard for mutation testing is based on the College of American Pathologists (CAP) and International Association of Study of Lung Cancer (IASLC) guidelines, adopted by American Society of Clinical Oncology (ASCO) in 2014 and then revised in 2018. This recommendation is also currently supported by NCCN guidelines. However, the molecular testing space continues to evolve rapidly, given the improving technology applications for both tissue and blood.

Tissue diagnosis

Acquiring adequate tissue at the time of diagnosis is of utmost importance for a complete diagnosis, including both histological identification and the successful completion of molecular testing. Standard analytic methods for molecular testing must be able to detect mutations in a sample with a minimum of 20% malignant tumor cells. There are many methods of tissue testing, including immunohistochemistry (IHC), fluorescent in situ hybridization (FISH), polymerase chain reaction (PCR), and next-generation sequencing (NGS). Each of these platforms provides specific information and has various advantages and disadvantages. The most used testing for EGFR mutation status and therefore selection of EGFR TKI therapy includes PCR-based and/or NGS-based panels. Defined by the CAP/IASLC guidelines as a strong recommendation, IHC testing of EGFR protein expression is not appropriate for selection of EGFR TKI therapy, as EGFR protein overexpression has not been shown to be associated with EGFR TKI sensitivity. Data from tissue IHC studies of specific EGFR variants (e.g., exon 21 L858R, exon 19 deletion E746-A750del) have shown reasonable sensitivity and specificity. For example, in a sample of 79 EGFR mutation−positive and 29 EGFR mutation−negative cases (diagnosed via reflex PCR testing), the overall sensitivity for IHC was 84.8%, with a specificity of 100%, when compared with the PCR-based method. However, these IHC-based assays are limited due to the challenges in multiplexing (i.e., testing for many gene variants at a single time). The CAP/IASLC guidelines also do not recommend EGFR copy number analysis by FISH for selection of EGFR TKI therapy, as copy number amplifications in EGFR have not been shown to correlate with response to EGFR TKI therapy.

As previously mentioned, one of the main diagnostic methods used for EGFR mutation testing is PCR-based assays, which can detect predefined specific variants such as point mutations. PCR amplifies DNA sequences of interest, allowing for a relatively small amount of tissue required for analysis than other methods and a relatively rapid turnaround time. There is currently a US Food and Drug Administration (FDA)-approved PCR-based test for EGFR mutations for both tissue and plasma, cobas EGFR mutation test version 2, which can detect 42 mutations in exons 18, 19, 20, and 21 of the EGFR gene including the T790M resistance mutation. Finally, NGS is emerging as the standard of care upfront molecular testing, as it is a highly multiplexed technology and assesses the range of mutation variants, copy number changes, and rearrangements/fusions. At least three NGS-based assays have been approved by the FDA for marketing authorization, and many additional commercial-targeted NGS assays are available. The CAP/IASLC guidelines also recommend, on the basis of expert consensus, that multiplexed genetic sequencing panels are preferred where available over multiple single-gene tests. However, the cost is more expensive and turnaround time can be longer for NGS panels. Therefore there may still be a role for performing concurrently (where permitted) the more “rapid” PCR-based EGFR tissue test or a liquid biopsy (discussed later), particularly if there is a high suspicion for an EGFR mutation based on the patient’s demographics and clinical presentation.

Liquid biopsies and circulating DNA

The diagnostic paradigm in NSCLC has opened the door for liquid biopsies or plasma-based molecular testing as part of routine care. Even in the era of plasma-based molecular testing, tissue-based molecular analysis is recommended when the plasma test is negative given the imperfect sensitivity of the available assays . Liquid biopsies involve isolating and analyzing circulating tumor/cell-free DNA (ct/cfDNA) and are an increasingly powerful tool for noninvasive molecular testing, with many studies exploring the diagnostic clinical utility cfDNA. There are different methods of cfDNA testing including allele-specific PCR-based assays, digital PCR, BEAMing (PCR and flow cytometry), and a variety of NGS methods (i.e., amplicon-based, capture-based, whole exome, and whole genome). These methods vary on their detection limit, with capture-based NGS having the highest sensitivity currently, and their ability to detect predefined (e.g., PCR-based) versus broader variants (e.g., NGS mutations, indels, copy number variations, genomic rearrangements). The IASLC has also released a 2018 statement paper on liquid biopsy testing, supporting validated PCR-based methods as appropriate for evaluating predefined specific mutations, such as EGFR, and stating that more comprehensive NGS assays are also an option (where available) and can even be considered when concurrent tissue molecular analysis will take longer than 2 weeks to result.

The cobas EGFR mutation test version 2, a real-time quantitative PCR-based assay, is currently the only FDA-approved liquid biopsy test for NSCLC, although other countries may have different approvals. An analysis of three studies of frontline erlotinib versus chemotherapy (ENSURE, FASTACT-2, ASPIRATION) demonstrated this liquid biopsy test, when compared with tissue cobas EGFR mutation test v1, had a concordance rate of 84.4%, sensitivity of 72.1% (95% CI, 67.8–76.1), and specificity of 97.9% (95% CI, 96.0–99.0), with sensitivity being greatest in patients with higher baseline tumor burden. Real-world data as demonstrated in the ASSESS trial of 1288 European and Japanese patients had a low sensitivity rate of 46% despite a high concordance rate and specificity, although with the limitation of variable PCR-based and sequencing methods used for the comparison of cfDNA analysis. Despite its approval, the applicability of cobas EGFR mutation test version 2 is in question given available assays with higher sensitivity for detection of EGFR mutations. As an example of the improved sensitivity of NGS-based plasma cfDNA test, the NILE study prospectively examined patients with treatment-naïve advanced NSCLC who underwent a validated NGS-based cfDNA test (Guradant360) and compared it with standard of care tissue genotyping (by physician discretion, allowed for NGS, PCR, FISH, IHC, or Sanger sequencing). Of the 282 patients in this study, when examining the EGFR mutation subset (tissue positive n = 32), cfDNA testing for EGFR exon 19 deletions had a concordance rate of 98.2%, sensitivity of 81.8%, and a specificity of 100%, while the analysis for cfDNA testing for EGFR exon 21 L858R mutations was 99.6%, 90%, and 100%, respectively.

In summary, plasma cfDNA may be part of the standard of care workup for patients with EGFR -mutated NSCLC, particularly if there is insufficient tissue for molecular testing or if the turnaround time of plasma-based molecular testing is predicted to be faster than the tissue-based molecular testing. It is important to note that if there is a negative test result for EGFR mutation on plasma testing, a rebiopsy for tissue analysis is recommended given the imperfect sensitivity. On the opposite end, a positive result for an EGFR mutation on plasma testing is enough evidence for treatment decision making. Liquid biopsies’ role may evolve to include more than upfront initial diagnostics. For example, analysis of plasma cfDNA may allow for routine clinical monitoring, as an addition to surveillance imaging, as levels can decrease during treatment in conjunction with response and increase in advance of disease progression. In addition, analysis of plasma cfDNA may inform prognosis with treatment, as the complete clearance of mutations from plasma cfDNA has been associated with improved survival outcomes. The role for liquid biopsy at the time of resistance to EGFR TKI therapy is discussed in further detail as follows.

Overview of generation of EGFR tyrosine kinase inhibitors

As described in Table 10.1 , there are currently three generations of EGFR TKIs available for the treatment of frontline advanced EGFR -mutated NSCLC. The first generation of EGFR TKIs include erlotinib, gefitinib, and icotinib, which bind the EGFR tyrosine kinase domain in a reversible manner and also target wild-type EGFR . Icotinib is not available in the United States, only in China. The second-generation EGFR TKIs include afatinib and dacomitinib, which irreversibly inhibit the pan ErbB/HER family (including EGFR) and have significant EGFR wild-type activity. All patients receiving EGFR TKIs eventually develop resistance, and one of the most common mechanisms of resistance to first- and second-generation EGFR TKIs includes development of the gatekeeper mutation T790M, which increases receptor affinity for adenosine-5-triphosphate and results in sterical blockade of both first- and second-generation EGFR TKIs.

Osimertinib is a third-generation EGFR TKI that binds irreversibly, forming a covalent bond at the cysteine residue of position 797 of the tyrosine kinase domain and, importantly, targeting both sensitizing EGFR mutations and T790M along with being relatively sparing of EGFR wild type. The relative binding to EGFR wild type is important across the different generations of EGFR TKIs because on target EGFR wild-type effects result in skin and GI toxicity in an often dose-dependent manner. Osimertinib’s first approval was in the second-line setting after progression on treatment with first-line EGFR TKI and in tumors harboring the EGFR T790M mutation ; however, osimertinib has now moved to the frontline setting. ^,

Current treatment paradigm

It is clear that EGFR TKIs improve PFS compared with platinum-based chemotherapy and that third-generation EGFR TKIs (i.e., osimertinib) outperform first-generation EGFR TKIs (i.e., erlotinib and gefitinib), with the latter demonstrating an emerging signal of improved OS. In a large meta-analysis from 2013 of 23 trials (13 frontline, 7 second-line, and 3 maintenance; n = 14,570) with EGFR mutation status known in 31% of patients, EGFR TKIs significantly prolonged PFS in the frontline setting, in patients with EGFR mutation−positive disease (HR, 0.43; 95% CI, 0.38−0.49; P < 0.01), while no effect on survival was observed. This meta-analysis included most of the frontline randomized trials at that time that compared EGFR-TKIs (i.e., gefitinib, erlotinib, afatinib) to platinum doublet chemotherapy, also summarized in Table 10.2 . Since this meta-analysis, there have been, of course, updated data of prior studies, along with additional data supporting the use of new EGFR TKIs including second-generation dacomitinib and third-generation osimertinib, ^, with both of these studies demonstrating improvement in PFS and OS compared with first-generation EGFR TKIs; however, there were critical differences in trial design, which are addressed in further detail later. Box 10.1 highlights the overall goals of therapy in the treatment of metastatic EGFR- mutated NSCLC.

TABLE 10.2 ■

Key Trials for First- and Second-Generation Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitors

Trial (Year)	N	TKI	Comparison Arm	ORR (%)	Median PFS (Months)	Median OS (Months)
IPASS ^, (2009)	EGFR subset: 261	Gefitinib	Carboplatin/Paclitaxel	71.2% vs. 47.3%	9.5 vs. 6.3 (HR, 0.48; 95% CI, 0.36–0.64; P < 0.001)	21.6 vs. 21.9 (HR, 1.0) NS Data at 78% maturity
NEJ002 ^, (2010)	230	Gefitinib	Carboplatin/Paclitaxel	73.7% vs. 30.7%	10.8 vs. 5.4 (HR, 0.30; 95% CI, 0.22–0.41; P < 0.001)	27.7 vs. 26.6 (HR, 0.89) NS
WJTOG 3405 (2010)	172	Gefitinib	Cisplatin/Docetaxel	62.1% vs. 32.2%	9.2 vs. 6.3 (HR, 0.49; 95% CI, 0.34–0.71; P < 0.001)	34.8 vs. 37.3 (HR, 1.25) NS
OPTIMAL (2011)	165	Erlotinib	Carboplatin/Gemcitabine	83% vs. 36%	13.1 vs. 4.6 (HR, 0.16; 95% CI, 0.10–0.26; P < 0.0001)	22.8 vs. 27.2 (HR, 1.19) NS
EURTAC (2012)	174	Erlotinib	Cisplatin/Docetaxel or Gemcitabine (Carboplatin allowed as alternative to cisplatin)	64% vs. 18%	9.7 vs. 5.2 (HR, 0.37; 95% CI, 0.25–0.54; P < 0.0001)	19.3 vs. 19.5 (HR, 1.04), NS
LUX-Lung 3 (2013)	345	Afatinib	Cisplatin/Pemetrexed	56% vs. 23%	11.1 vs. 6.9 ^a (HR, 0.58; 95% CI, 0.43–0.78; P = 0.0004) Del19/L858R 13.6 vs. 16.9 (HR, 0.47; 95% CI, 0.34–0.65; P < 0.0001)	28.2 vs. 28.2 (HR, 0.88) NS
LUX-Lung 6 (2014)	364	Afatinib	Cisplatin/Gemcitabine	67% vs. 23%	11 vs. 5.6 (HR, 0.28; 95% CI, 0.20–0.39; P < 0.0001)	23.1 vs. 23.5 (HR, 0.93) NS
Inter-TKI Comparisons
WJOG 5108L (2016) ^b	561	Gefitinib	Erlotinib	45.9% vs. 44.1%	6.5 vs. 7.5 (HR, 1.125; 95% CI, 0.94–1.35; P = 0.257)	22.8 vs. 24.5 (HR, 1.03) NS
LUX-Lung 7 ^, (2016)	319	Afatinib	Gefitinib	72.5% vs. 56%	11 vs. 10.9 (HR, 0.74; 95% CI, 0.57–0.95; P = 0.012)	27.9 vs. 24.5 (HR, 0.86) NS
ARCHER 1050 (2018)	452	Dacomitinib	Gefitinib	75% vs. 72%	14.7 vs. 9.2 (HR, 0.59; 95% CI, 0.47–0.74; P < 0.0001)	34.1 vs. 26.8 (HR, 0.76; P = 0.044) statistically significant

All primary endpoints were progression-free survival (PFS), with hazard ratios (HR), 95% confidence intervals (CI), and P values presented above. Data are presented in the objective response rates (ORR), median PFS, and median overall survival (OS) columns as: result of TKI vs. comparison arm. All these trials included EGFR exon 19 deletions and exon 21 L858R mutations unless otherwise noted.

EGFR, Epidermal growth factor receptor; NS, not statistically significant; TKI, tyrosine kinase inhibitor.

a LUX-Lung-3 included 11% mutations other than exon 19 deletion and exon 21 L858R mutations. Both total group and EGFR exon 19 deletion/exon 21 L858R PFS are presented above.

b WJOG 5108 did include 16%–18% EGFR wild type.

BOX 10.1 ■

Goals of Treatment

■
Use the “best” drugs first (when available) with documentation of improvement in overall survival being the gold standard.
■
Be cautious in intertrial comparisons of progression-free survival (PFS) in randomized epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI) studies.
■
Improve quality of life, minimize toxicity.

First-line therapy for metastatic EGFR -mutated NSCLC

Although there are several EGFR TKIs available for use in the United States, in analyzing the available data, the preferred first-line standard of care treatment of advanced EGFR mutant NSCLC is with third-generation EGFR TKI, osimertinib . Before osimertinib moving to the frontline based on the FLAURA data, the AURA3 clinical trial showed a significant improvement in PFS with osimertinib compared with platinum-pemetrexed–based chemotherapy for those patients who progressed and developed the EGFR gatekeeper resistance mutation T790M on first- and second-generation EGFR TKIs.

However, the paradigm has shifted, as osimertinib has moved to the frontline given superior PFS and OS data compared with first-generation EGFR TKIs gefitinib and erlotinib, along with superior central nervous system (CNS) activity. Initial results of promising activity of osimertinib in EGFR TKI–naïve patients emerged early during the AURA phase I study dose expansion cohort of 60 patients, examining at osimertinib 80 mg and 160 mg daily dosing, with updated results (2019) reporting an ORR of 67% and 87% (77% overall). The median DOR was 19.3 months and 16.7 months, and a median PFS was 22.1 months and 20.5 months, for the 80 mg and 160 mg doses, respectively. The results from the phase III randomized FLAURA trial confirmed these findings and resulted in FDA approval of frontline osimertinib for advanced EGFR -mutated (exon 19 deletion or exon 21 L858R) NSCLC. This study included 556 patients randomized in a 1:1 fashion to standard dosing osimertinib 80 mg daily ( n = 279) or erlotinib 150 mg daily or gefitinib 250 mg daily ( n = 277, 66% gefitinib and 34% erlotinib). Patients with neurologically stable CNS metastases were eligible, including 19% in the osimertinib arm and 23% in the erlotinib/gefitinib arm. The primary endpoint was achieved favoring osimertinib over gefitinib/erlotinib, translating into a median PFS 18.9 months versus 10.2 months, respectively (HR, 0.46; 95% CI, 0.37–0.57; P < 0.001). All predefined subgroups examined clearly favored treatment with osimertinib. Given the high response rates observed even with earlier-generation EGFR TKIs, the ORR was not significantly different between the two treatment arms (80% osimertinib arm vs. 76% erlotinib/gefitinib arm). However, the median duration of response was longer with osimertinib (17.2 months vs. 8.5 months). Highlighting the CNS activity of osimertinib, a similar PFS benefit favoring osimertinib over gefitinib/erlotinib was observed in those with and without baseline CNS metastasis (HR, 0.47; 95% CI, 0.30–0.74; P <0.001 and HR, 0.46; 95% CI, 0.36–0.59; P < 0.001, respectively). Overall, the rates of CNS progression events were also lower in the osimertinib arm compared with the gefitinib/erlotinib arm (6%, 17 patients vs. 15%, 42 patients, respectively). In those with CNS lesions, the median CNS PFS was not reached with osimertinib and 13.9 months in the gefitinib/erlotinib arm (HR, 0.48; 95% CI, 0.26–0.86; P = 0.014). In interpreting these results, it is important to note that baseline brain imaging was not mandated in the FLAURA study, which is otherwise a standard for staging for NSCLC, and only patients with confirmed CNS metastases were followed on study with brain imaging.

Not surprisingly, given the differences in mechanisms of action of third-generation EGFR TKIs versus first-generation EGFR TKIs, there was less rash/acne observed in the osimertinib group compared with the gefitinib/erlotinib group (58% vs. 78%), although the rates of the following adverse events were similar including diarrhea (50% vs. 57%), dry skin (36% in both arms), paronychia (35% vs. 33%), and stomatitis (29% vs. 20%). There are currently no data comparing third-generation EGFR TKIs to second-generation agents such as afatinib or dacomitinib for efficacy.

The proportion of adverse events leading to permanent discontinuation were also slightly lower in the osimertinib group when compared with the erlotinib/gefitinib group (13%, 37 patients vs. 18%, 49 patients, respectively). Despite these events being relatively rare, osimertinib does have a unique cardiotoxicity profile including prolongation of the QT interval and depressed left ventricular ejection fraction, which may require additional cardiac surveillance with electrocardiograms and transthoracic echocardiograms, respectively, particularly in patients who have a cardiac history.

At the time of the initial publication of the FLAURA study, the data on the secondary endpoint of OS were immature (25% maturity). However, these results were recently reported at 58% maturity of the OS data, with the OS observed favoring osimertinib and being statistically significant with the upper bounds of the 95% CI approaching but not crossing 1: median OS of 38.6 months osimertinib versus 31.7 months erlotinib/gefitinib (HR, 0.799; 95% CI, 0.647–0.997; P = 0.0462) ( Fig. 10.1 ) . Given the less robust OS hazard ratio, not all predefined subgroups examined demonstrated a clear OS benefit with osimertinib over the comparator arm with the hazard ratio either at or approaching 1.0 (e.g., Asian race, EGFR exon 21 L858R mutation). In interpretation of OS data, it is always important to note any discrepancies in subsequent therapies between the two treatment arms. In the osimertinib arm, 48% (133 patients) of patients randomized went on to start a first subsequent therapy (68% chemotherapy and 29% EGFR TKI other than osimertinib) and 22% remained on osimertinib at the time of analysis. In the comparator EGFR TKI arm, 65% (180 patients) of patients randomized went on to start a first subsequent therapy (22% chemotherapy, 47% osimertinib, and 27% EGFR TKI other than osimertinib) and only 5% remained on the trial EGFR TKI at the time of analysis. Despite one third of patients receiving another EGFR TKI after osimertinib as their first subsequent therapy, there are no data to suggest this would be a beneficial strategy; however, the OS benefit remained.

Fig. 10.1, FLAURA: progression-free survival (PFS) and overall survival (OS) with osimertinib in frontline epidermal growth factor receptor (EGFR)-mutated non−small cell lung cancer (NSCLC).

In addition, the FLAURA trial had a crossover rate of 31% (85 of 277) from the erlotinib/gefitinib arm to the osimertinib arm, which also strengthens the OS findings. However, the EGFR T790M resistance mutation rate is expected to be ∼50% after progression on first-generation EGFR TKIs, and therefore a subset of patients likely eligible to receive osimertinib second line did not do so. The OS data also emphasize the principle to use the best therapy first for our patients, as 22% of patients randomized in both the osimertinib and erlotinib/gefitinib arms did not go on to first subsequent therapy due to death. In summary, the OS data from FLAURA are particularly striking, as crossover has been the major cited reason for the lack of OS benefit seen in prior EGFR TKI versus platinum-doublet chemotherapy studies.

Since it has already been established from multiple randomized phase 3 studies (see Table 10.2 ) that EGFR TKIs (i.e., erlotinib, gefitinib, afatinib) compared with platinum doublet chemotherapy are superior, if osimertinib is not available, there are other inter-EGFR TKI trials examining first- versus second-generation EGFR TKIs to potentially guide choice of frontline therapy. However, there are issues with cross-trial comparisons, which can be misleading given heterogeneity in patient populations and differences in clinical trial designs.

Some data suggest that second-generation EGFR TKIs may be better than first-generation EGFR TKIs but at a higher toxicity cost. In the Phase IIb LUX-Lung 7 study, 319 treatment-naïve patients with EGFR -mutated NSCLC were randomly assigned to afatinib or gefitinib, with afatinib showing a small but statistically significant PFS benefit (11 vs. 10.9 months; HR, 0.73; 95% CI, 0.57–0.95; P = 0.017) and time to treatment failure (13.7 vs. 11.5 months). While initial data reporting had immature OS data, after a median follow-up of 42.6 months, median OS with afatinib was 27.9 months versus 24.5 months with afatinib (HR, 0.86; 95% CI, 0.66–1.12; P = 0.26) and was not statistically significant. Regardless, these analyses were exploratory and not preplanned. The modest benefit, the lack of a predefined statistical plan, and the higher grade 3 or 4 rates of diarrhea and rash/acne in the afatinib group place the applicability of these results (i.e., using second-generation EGFR TKI over first-generation EGFR TKI) in clinical practice in question.

With a more robust trial design, ARCHER 1050 was a randomized open-label phase III study examining second-generation TKI dacomitinib compared with gefitinib in 452 patients. In the 452 patients with EGFR- mutated NSCLC randomized in this study, the dacomitinib group had an improved PFS of 14.7 months versus 9.2 months (HR, 0.59; 95% CI, 0.47–0.74; P < 0.0001) and therefore met the statistical plan target PFS HR of ≤0.667. The median OS was also improved in the dacomitinib arm at 34 months versus 26.8 months in the gefitinib arm at median follow-up of 31 months (HR, 0.76; 95% CI, 0.582–0.993; P = 0.0438). Recently, as of ESMO Asia 2019, these OS results remained significant with ongoing follow-up, even for patients who required dose reductions, with an updated OS benefit of 37.7 months. Of note, an unusual eligibility criterion for the study was exclusion of patients with any history of brain metastases, which is a key difference compared with other frontline trials and may have impacted the OS results observed. The rationale for this per the study investigators was that at the time of the study design, the impact of dacomitinib in the CNS was unknown and gefitinib only has relative activity in the CNS. This significant difference in the patient population, the safety profile of dacomitinib including a higher rate of grade 3/4 dermatitis acneiform and diarrhea along with a higher rate of dose interruptions (78% vs. 54%), and the fact that this drug was the latest EGFR TKI to gain approval will likely limit its use in the United States.

Summary of frontline treatment of metastatic EGFR -mutated NSCLC

Fig. 10.2 shows the potential options for frontline EGFR TKI treatment with currently available data. At this time, osimertinib has the best frontline data and should be used frontline where available. Some argue that summing the PFS of first- or second-generation EGFR TKI followed by osimertinib could potentially result in longer PFS compared with using osimertinib frontline. However, we strongly caution against the approach of summing the PFS benefit of different trials and recommend using the best drug first, given two key considerations :

1.
Many patients with advanced NSCLC do not receive second-line therapy . Trials in NSCLC have reported anywhere from 30% to 70% of patients receiving second-line therapy. Since a subset of patients do not move on to second-line therapy, due to disease progression, changes in performance status, or other clinical changes, it is important to use the best treatment first.
2.
Not all patients will develop a T790M mutation as part of the resistance mechanism to first- and second-generation EGFR TKIs, therefore minimizing the utility of osimertinib in the second-line setting. There is currently no prediction model for the development of T790M mutation, which occurs in 50%–60% of patients after progression on first- or second-generation EGFR TKI, and the second-line indication of the use of osimertinib is restricted to inclusion of this mutation.

Fig. 10.2, First-line options for epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI) therapy.

However, we understand that one must consider cost and access to frontline treatments. In resource-limited settings, the sequencing of first- and second-generation EGFR TKIs followed by third-generation EGFR TKIs or chemotherapy may be required, depending on whether third-generation EGFR TKI trials are available. In addition, the combination of earlier-generation EGFR TKIs with chemotherapy may be an option, as is discussed further later, although this requires patients to be fit with an excellent performance status, given the cumulative adverse events and symptoms possible with combination treatments.

Use of chemotherapy in the first- and second-line settings

There is an interest in using chemotherapy in combination with EGFR TKIs frontline to prevent resistance and prolong disease control. Using both together, particularly if an OS benefit can be demonstrated, eliminates the concern of a subset of patients not being able to receive subsequent chemotherapy due to disease progression and/or clinical decline. Recently, two randomized phase III studies have assessed the combination of concurrent gefitinib with carboplatin/pemetrexed chemotherapy upfront, followed by gefitinib and pemetrexed maintenance. These studies were inspired by the initial positive results of the NEJ005 study showing an improvement in median OS with concurrent gefitinib and chemotherapy when compared with alternating gefitinib and chemotherapy (median OS of 41.9 vs. 30.7 months, respectively; P = 0.036). The Japanese phase III NEJ009 trial randomized 345 patients with newly diagnosed EGFR -mutated NSCLC to gefitinib plus carboplatin and pemetrexed versus gefitinib alone, followed by platinum-based regimen at progression, with a primary endpoint of PFS, followed by PFS2 and OS using hierarchical testing. The combination group showed an improvement in ORR (84% vs. 67%, P < 0.001), median PFS (20.9 months vs. 11.2 months; HR, 0.49; 95% CI, 0.39–0.62; P < 0.001), and median OS (50.9 months vs. 38.8 months; HR, 0.72; 95% CI, 0.55–0.95; P = 0.021). The positive OS benefit was considered exploratory due to the hierarchical testing method because there was no difference between the treatment arms in the predefined PFS2 endpoint, defined as the period from the date of randomization until both platinum-based therapy and gefitinib became ineffective “conceptually.” In the study, only 22% of patients received osimertinib after gefitinib use, which is lower than expected given the known rates of T790M. Not surprisingly, there was higher toxicity, mainly hematological, when combining chemotherapy and gefitinib (65% grade 3 or higher treatment-related AEs in the combination arm vs. 31% in the gefitinib arm). Despite this, the study did not find significant differences in global quality-of-life scores.

A similar phase III randomized trial design of gefitinib with or without carboplatin/pemetrexed chemotherapy was also recently reported with the Indian experience, with a total of 350 patients included. The estimated median PFS of 16 months in the combination arm was superior to the gefitinib alone arm of 8 months (HR, 0.51; 95% CI, 0.39–0.66; P < 0.001). The secondary endpoint of OS was also prolonged in the combination arm (not reached vs. 17 months; HR, 0.45; 95% CI, 0.31–0.65; P < 0.001). Only a small proportion of patients received osimertinib at disease progression, including 11% in the combination arm and 15% in the gefitinib alone arm. There was also a higher rate of grade 3 or greater adverse events in the chemotherapy plus gefitinib arm compared with the gefitinib alone arm (51% vs. 25.3%, respectively), again largely hematological. There was also a notable 16% pemetrexed discontinuation rate due to toxicity, including nephrotoxicity in 11%. Quality-of-life data have been collected but have not been reported.

There were several differences in the population characteristics (e.g., median age, proportion of male sex, smoking status) between the Indian versus Japanese study, with the most important difference being the inclusion of patients with ECOG performance status of 2 (21% of patients) in the Indian study. The inclusion of these patients provided applicability of this strategy to a more generalizable population and also potentially explained the numerically lower numbers of PFS and OS observed in the Indian study. Both studies included patients with brain metastasis and included uncommon sensitizing EGFR mutations beyond exon 19 deletion and exon 21 L858R. The data from these studies of combination chemotherapy with first-generation EGFR TKIs provide compelling data for alternative frontline treatments when osimertinib is unavailable, with comparable PFS and OS outcomes. It is important to note that in these studies, a small subset of patients went on to receive third-generation EGFR TKIs at progression, which is an important option for patients who develop T790M resistance. Given the promising, even practice-changing data above with earlier-generation EGFR TKIs, there is a planned phase III trial (FLAURA2) examining osimertinib with platinum/pemetrexed chemotherapy compared with osimertinib alone, with the primary endpoint of PFS.

The data from frontline combination chemotherapy and EGFR TKI differ from the phase III IMPRESS trial, conducted in the second-line setting, with 265 chemotherapy-naïve patients with EGFR -mutated advanced NSCLC who had progressed on gefitinib. The trial assessed the question of whether continuing EGFR TKI therapy with chemotherapy was beneficial at the time of progression from an EGFR TKI (i.e., continuing gefitinib with cisplatin and pemetrexed vs. placebo plus chemotherapy). In the intention-to-treat population, there was no significant difference in median PFS (5.4 months in both treatment arms), with updated OS data showing a surprising detrimental effect of continuing the TKI with chemotherapy at progression (median OS 13.4 months vs. 19.5 months; HR, 1.44; 95% CI, 1.07–1.94; P = 0.016). This difference was statistically significant in the patients with EGFR T790M mutation–positive plasma samples, whereas statistical significance was not reached in the patients with T790M mutation–negative plasma samples. However, more patients in the placebo group were exposed to subsequent treatment at the time of progression compared with the gefitinib group. Furthermore, there were more patients with CNS metastasis at baseline in the gefitinib group compared with the placebo group, and they were not adjusted for OS results. In summary, there is a potential role for combination upfront chemotherapy plus EGFR TKIs, particularly first-generation EGFR TKIs where third-generation EGFR TKIs are not available; however, further studies are needed looking at continuing EGFR TKI post progression with chemotherapy and could potentially be detrimental.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here