Introduction

Epidermal growth factor receptor (EGFR) mutations are one of the most common oncogenic driver mutations, occurring in 10% to 20% of Caucasian patients and approximately 50% of Asian patients with non-small cell lung cancer (NSCLC).[1] With the development of sensitive detection technologies such as next-generation sequencing-based repeated deep sequencing, many patients have been found to harbor different mutations in addition to a single EGFR mutation.[2] There are two main types of additional mutations: independent mutations (complex/double/multiple) in the EGFR-tyrosine kinase domain (EGFR compound mutations) and additional mutations in other pathways (EGFR concomitant mutations) [Figure 1].[3] While extensive heterogeneity was found among different EGFR compound and concomitant mutations, most such studies were case reports or small-scale retrospective studies, and the therapeutic efficacy remains unexplored. Therefore, we aimed to review current studies on both EGFR compound mutations and EGFR concomitant mutations to explore their prognostic impact in patients with NSCLC and recommend optimal treatment strategies.

Figure 1:

Common forms of EGFR mutations. EGFR: Epidermal growth factor receptor; ex19del: Exon 19 in-frame deletion; ex20ins: Exon 20 insertion; BRAF: B-Raf proto-oncogene, serine/threonine kinase; CCNE1: Cyclin E1; CDK4: Cyclin dependent kinase 4; CDK6: Cyclin dependent kinase 6; CTNNB1: Catenin beta 1; MET: Mesenchymal-epithelial transition factor; MYC: Myelocytomatosis; NKX2-1: NK2 homeobox 1; PIK3CA: Phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha; RB1: RB transcriptional corepressor 1; TP53: Tumor protein p53.

Different Sensitivities of EGFR Mutations to EGFR-Tyrosine Kinase Inhibitors (TKIs)

Currently, the most common types of EGFR mutations are typical activating mutations, including exon 19 in-frame deletion (ex19del) around the long-range epigenetically active motif (about 45% of EGFR mutations), followed by the L858R point mutation in exon 21 (approximately 40% of EGFR mutations).[2] Other atypical mutations include exon 18 mutations (3%–5%, including G719X, E709X, and Del18) and exon 20 mutations (1%–5%, including exon 21 insertion [ex20ins] and S768I).[2,3] Genotyping tumors, stratifying patients per different mutation types, and administering appropriate targeted agents could improve clinical outcomes of patients with EGFR-mutant NSCLC.

EGFR-TKIs are the primary treatment option for patients with EGFR mutation-positive NSCLC, and their efficacy has been reported for patients with advanced and early stages in adjuvant settings.[4–6] The EGFR-TKIs used in clinical practice include first-generation EGFR-TKIs (gefitinib, erlotinib, and icotinib), second-generation EGFR-TKIs (afatinib and dacomitinib), and third-generation EGFR-TKIs (osimertinib, aumolertinib, furmonertinib, and lazertinib). However, the sensitivity of different EGFR mutations to EGFR-TKIs varies. Patients with ex19del and L858R achieved an objective response rate (ORR) of over 70% and 60%, respectively, to different generations of EGFR-TKIs. Meanwhile, patients with rare mutations such as G719X and L861Q could achieve an ORR of approximately 50%, and patients with exon 20 mutations hardly responded to first- and second-generation TKIs (ORR <10%). Notably, patients with the T790M mutation achieved an ORR over 60% to third-generation EGFR-TKIs, while most patients with other exon 20 mutations, including ex20ins, performed poorly after receiving third-generation TKIs. However, heterogeneity was found between studies, and the reported ORR range was 0 to 60%[7–12].

EGFR Compound Mutations

EGFR compound mutations are defined as complex, double, or multiple independent mutations in the EGFR-tyrosine kinase domain.[3] The frequency of EGFR compound mutations varies among studies and populations, with previous studies indicating an incidence rate of approximately 4% to 26% among Asians and 5% to 7% among Caucasians with EGFR mutation-positive NSCLC.[13]EGFR compound mutations include three major types: combination of atypical mutations with typical mutations (approximately 50%), compound atypical mutations, and compound typical mutations.[14] In addition, the incidence of the different mutations also varies; for example, G719X was found in >90% of complex EGFR mutations.[11]

EGFR compound mutations are associated with poor prognosis

Various EGFR compound mutations can alter the tertiary structure of EGFR proteins, and this affects sensitivity to EGFR-TKIs and prognosis of patients with NSCLC. A previous study explored the prognosis of patients with lung adenocarcinoma who underwent curative resection and received adjuvant chemotherapy. Patients with simple EGFR mutations (89.1% harboring typical mutations) had a similar disease-free survival to patients with compound EGFR mutations but a superior overall survival (OS) (83.7 months vs. 72.8 months, P = 0.02) than those with compound EGFR mutations. Patients with compound EGFR mutations also have higher burdens of missense mutations and co-alterations in other genes.[2] A retrospective observational study explored the survival analysis of patients with metastatic NSCLC treated with first- or second-generation EGFR-TKIs; patients with EGFR compound mutations had a superior median OS (31.0 [95% confidence interval : 18.2–43.7] months vs. 17.0 [95% CI: 8.2–25.7] months, P = 0.04) and a tendency toward improved median progression-free survival (PFS) (12.3 [95% CI: 8.5–15.4] months vs. 8.3 [95% CI: 4.8–11.7] months, P = 0.06) than patients harboring EGFR exon 18 mutations.[15] Another retrospective multicenter study NCT04179890 also reported that patients with compound mutations who received EGFR-TKIs had more favorable outcomes (median time-to-treatment failure: 12.3 months, 95% CI: 8.5–15.5 months; median OS: 28.7 months, 95% CI: 2.5–33.0 months, ORR: 48.6%) than patients harboring T790M and ex20ins mutations.[16] Thus, patients with compound EGFR mutations seem to have poorer prognoses than patients with simple mutations who are sensitive to first-/second-generation EGFR-TKIs (ex19del, L858R), but a better prognosis than patients with simple mutations (such as T790M, ex20ins, and exon 18 mutations) who are insensitive to first-/second-generation EGFR-TKIs.

The types of compound mutation also affect the prognosis of patients with NSCLC. Patients with typical mutations alone showed superior median PFS (median PFS: 11.9 months, 95% CI: 10.4–13.5 months) than patients with both typical and atypical mutations (median PFS: 8.1 months, 95% CI: 3.2–13.1 months), patients with both typical mutations and T790M (median PFS: 8.0 months, 95% CI: 0.8–15.2 months), and patients with atypical mutations alone or complex atypical mutations (median PFS: 1.4 months, 95% CI: 0.3–2.5 months) after being treated with either gefitinib or erlotinib. Besides, patients with typical mutations alone also showed better treatment ORR (ORR: 74.8%) than patients with both typical and atypical mutations (ORR: 68.8%) and patients with atypical mutations alone or complex atypical mutations (ORR: 25.0%) (P <0.01).[17] A retrospective study reported the prognosis of patients with advanced EGFR compound mutations on first-generation EGFR-TKIs treatment. Notably, the ORR for patients with compound typical mutations, atypical mutation plus typical mutation, compound atypical mutations, and compound mutations harboring T790M or ex20ins was 75.0%, 60.0%, 71.0%, and 8.3%, respectively, and median PFS was 18.2 (95% CI: 10.6–25.9) months, 9.7 (95% CI: 3.3–15.8) months, 9.6 (95% CI: 3.3–19.0) months, and 1.4 (95% CI: 0.4–2.3) months, respectively.[18] Therefore, types of compound mutations affect sensitivity to EGFR-TKIs, which also depends on single mutations.[11]

Treatments for patients with EGFR compound mutations

Several studies have explored the efficacy of first-generation EGFR-TKIs in patients with EGFR mutations. Tu et al[19] indicated that patients with L858R compound mutations receiving first-generation EGFR-TKI treatment (gefitinib or erlotinib) had a median PFS of 15.2 (95% CI: 8.7–21.7) months and a median OS of 27.7 (95% CI: 21.8–33.6) months. A retrospective study reported that after receiving gefitinib or erlotinib, patients with EGFR single mutations performed better (ORR: 74.1%; median PFS: 8.5 months; median OS: 19.6 months) than patients with a combination of atypical and typical mutations (OR: 60.0%; median PFS: 5.3 months; median OS: 18.8 months).[20] Another retrospective study showed that the efficacy of first-generation EGFR-TKIs in patients with a combination of atypical and typical mutations (median PFS: 9.8 months, 95% CI: 0.7–18.9 months; median OS: 21.1 months, 95% CI: 9.7–32.5 months) was equivalent to the efficacy in patients with compound typical mutations (median PFS: 9.5 months, 95% CI: 0–19.4 months; median OS: 23.3 months, 95% CI: 13.4–33.3 months) and was better than patients with the combination of typical and T790M mutations (median PFS: 1.9 months, 95% CI: 0–4.4 months; median OS: 16.9 months, 95% CI: 11.0–22.7 months).[21] Thus, first-generation EGFR-TKIs might be considered for patients with sensitive EGFR compound mutations but are less effective for patients with drug-resistant EGFR compound mutations.[21]

In addition, Kohsaka et al[11] explored the EGFR-TKI sensitivity of various EGFR compound mutations and found that the half maximal inhibitory concentration (IC50) of gefitinib in L858R_E709 compound mutations was 20 to 80 times higher than that in L858R single mutation. Despite compound mutations having less sensitivity to first-generation EGFR-TKIs, second-generation afatinib was effective in managing patients with compound mutations, with IC50 < 0.1 nmol/L. A pooled analysis with a database of 693 cases showed that treatment-naïve patients with EGFR compound mutations had an ORR of 77.1% and a duration of response (DoR) of 16.6 (95% CI: 13.8–18.7) months after afatinib treatment, whereas patients previously treated with EGFR-TKIs had an ORR of 28.6% and a DoR of 16.7 (95% CI: 9.9–21.8) months.[22] A retrospective multicenter study also showed that afatinib was effective in patients with compound mutations (median time of treatment: 12.3 months, 95% CI: 7.7–17.0 months; median OS: 29.1 months, 95% CI: 20.4–37.7 months) and in patients with major uncommon mutations (G719X and L861Q) (median time of treatment: 20.3 months, 95% CI: 15.1–25.5 months; median OS: 30.6 months, 95% CI: 26.3–34.8 months), but not in patients harboring drug-resistant mutations (ex20ins, S768I, T790M) (median time of treatment: 3.8 months, 95% CI: 1.7–6.0 months; median OS: 8.5 months, 95% CI: 5.2–11.7 months).[23] In addition, real-world data showed that afatinib reached an ORR of 100% in four patients with advanced NSCLC with exon 18 and 21 mutations.[24] Tan et al[25] retrospectively found that afatinib was superior to gefitinib/erlotinib/icotinib for patients with advanced or recurrent NSCLC with atypical plus typical mutations (ORR: 100.0% vs. 54.5%, P = 0.02; median PFS: not reached vs. 13.6 months, P = 0.03). These results indicated that afatinib could reduce heterogeneity and have a more extensive inhibitory profile than first- or third-generation TKIs, which supports afatinib as a treatment option for EGFR compound mutations.[26] Thus, afatinib could be recommended for complex mutations involving E709X, L861Q, or G719X, with the exception of drug-resistant mutations.[11,27] Further prospective clinical trials are required to confirm this hypothesis.

Furthermore, several studies have evaluated the efficacy of third-generation EGFR-TKIs in patients with EGFR mutations. In phase 2 of the KCSG-LU15-09 clinical trial, three of four patients with EGFR compound mutations (two patients with G719X + L861Q and two patients with S768I + G719X) received osimertinib and showed a reduction of over 30% in the target lesion size from that at baseline.[28] Another retrospective study found that the administration of osimertinib had superior survival benefit than first-generation EGFR-TKI for patients with ex20ins/T790M plus sensitive mutations (median PFS: 18.0 [95% CI: 15.1–20.9] months vs. 1.2 [95% CI: 0.9–1.6] months, P < 0.01; median OS: 25.1 months [95% CI: not calculable] vs. 17.3 [95% CI: 9.3–25.4] months, P = 0.02).[29] In addition, the ASTRIS study found that osimertinib showed better efficacy in T790M positive patients with advanced NSCLC with a combination of T790M and rare/complex mutations (T790M + G719X/S768I/ex20ins or T790M + two or more mutations) than patients in full analysis set (advanced T790M positive NSCLC patients who previously received EGFR-TKI) (ORR: 50.0% vs. 57.0%; median PFS: 8.1 months vs. 11.1 months).[30] In summary, third-generation EGFR-TKIs have clinical effects on EGFR compound mutations, especially complex mutations involving T790M, while the efficacy of third-generation EGFR-TKIs in patients with other compound exon 20 mutations should be further explored.

In general, treatment decisions for compound mutations are made as per specific mutation types. Although first-generation EGFR-TKIs are an effective treatment option for patients with EGFR compound mutations that contain sensitive mutations, afatinib is more sensitive and can reduce heterogeneity. Furthermore, for patients with T790M mutation, third-generation TKIs may be the optimal choice [Table 1].

Table 1 - Summary of the data and treatment options for patients with EGFR compound mutations. Treatment strategy Mutation type Clinical data References Treatment recommendation First-generation EGFR-TKIs L858R compound mutations Median PFS: 15.2 months; median OS: 27.7 months [19] First-generation EGFR-TKIs might be considered for patients with sensitive EGFR compound mutations Atypical and typical mutations ORR: 60%; median PFS: 5.3 months; median OS: 18.8 months [20] Atypical and typical mutations Median PFS: 9.8 months; median OS: 21.1 months [21] Compound typical mutations Median PFS: 9.5 months; median OS: 23.3 months Typical and T790M mutations Median PFS: 1.9 months; median OS: 16.9 months Atypical plus typical mutations ORR: 54.5%; median PFS: 13.6 months [25] ex20ins/T790M plus sensitive mutations Median PFS: 1.2 months; median OS: 17.3 months [29] Second-generation EGFR-TKIs Treatment-naive patients with EGFR compound mutations ORR: 77.1%; DoR: 16.6 months [22] Second-generation EGFR-TKI could be recommended for complex mutations involving E709X, L861Q, or G719X, except for drug-resistant mutations EGFR-TKI pretreated patients with EGFR compound mutations ORR: 28.6%; DoR: 16.7 months Compound mutations Median time of treatment: 12.3 months; median OS: 29.1 months [23] Exon 18 and 21 mutations ORR: 100% [24] Atypical plus typical mutation ORR: 100%; median PFS: NA [25] Third-generation EGFR-TKIs Compound mutation (two patients with G719X + L861Q; two patients with S768I + G719X) Over 30% reduction in the target lesion size from baseline [28] Third-generation EGFR-TKI had clinical effects on EGFR compound mutations, especially for complex mutations involving T790M ex20ins/T790M plus sensitive mutations Median PFS: 18.0 months; median OS: 25.1 months [29] T790M and rare/complex mutations (T790M + G719X/S768I/ex20ins or T790M + two or more mutations) ORR: 50.0%; median PFS: 8.1 months [30] T790M and common mutations ORR: 57.0%; median PFS: 11.1 months

DoR: Duration of response; EGFR-TKI: Epidermal growth factor receptor-tyrosine kinase inhibitor; ex20ins: Exon 20 insertion; NA: Not available; ORR: Objective response rate; OS: Overall survival; PFS: Progression-free survival.

EGFR Concomitant Mutations

When a patient harbors an EGFR mutation and mutations in other pathways, they are said to have concomitant EGFR mutations; 92.9% of patients with EGFR mutations harbor one or more concomitant mutations, and 89.9% of these concomitant mutations might have a functional impact.[31] Common types of concomitant EGFR mutations include tumor protein p53 (TP53), RB transcriptional corepressor 1 (RB1), catenin beta 1 (CTNNB1), phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA), B-Raf proto-oncogene, serine/threonine kinase (BRAF), mesenchymal-epithelial transition factor (MET), myelocytomatosis (MYC), cyclin dependent kinase 4 (CDK4), cyclin dependent kinase 6 (CDK6), NK2 homeobox 1 (NKX2-1), and cyclin E1 (CCNE1) targets.[31,32] Moreover, the incidence rate and type of concomitant mutations were not associated with EGFR mutation types but with the mean number of mutations and previous treatments.[33]

EGFR concomitant mutations are associated with poor prognosis

Different types of concomitant mutations have different effects on cancer cells. For example, patients with TP53 mutations have a higher somatic mutation burden and genomic instability. TP53 mutations might also allow tumor cells to evade cell cycle checkpoints and avoid apoptosis. Patients with TP53 and RB1 mutations have an increased risk of small cell lung cancer.[34–36] Mutations in PIK3CA and PTEN could activate the phosphatidylinositol 3 kinase (PI3K)/protein kinase B (AKT) pathway, and the CTNNB1 mutation could affect the Wnt/β-catenin pathway and result in tumor proliferation and progression.[37–40] Amplification of CDK4, CDK6, and CCNE1 affects the transition of the G1/S cell cycle.[41] Although other oncogenic driver gene mutations are considered mutually exclusive with EGFR mutations, EGFR concomitant driver alterations, including anaplastic lymphoma kinase (ALK) rearrangement, Kirsten rat sarcoma viral oncogene (KRAS) mutation, ROS proto-oncogene 1, receptor tyrosine kinase (ROS1) rearrangement, and MET alteration, might influence tumor cell proliferation, survival, invasion, and acquired resistance.[42] In addition, concomitant mutations could affect the tumor immune microenvironment. Zhang et al[43] showed that patients harboring EGFR concomitant mutations had a higher tumor mutational burden (TMB) than patients with EGFR mutations alone. Jin et al[44] found higher levels of programed cell death-ligand 1 (PD-L1) expression in patients with EGFR concomitant tumor suppressor gene mutations, such as TP53 and RB1, than in those with EGFR mutation alone. Moreover, patients with EGFR mutations had reduced natural killer CD56dim cells, cytotoxic cells, and T helper 2 cell infiltration in the tumor immune microenvironment when compared with patients with wild-type EGFR. Patients with EGFR-MAPK concomitant mutations had a similar TMB level, PD-L1 level, and immune microenvironment when compared with patients with wild-type EGFR.[45] The impact of different types of EGFR mutations on the tumor immune microenvironment warrants further investigation.

Concomitant mutations affect a patient's response to EGFR-TKI treatment, and prior studies have shown a higher incidence rate of concomitant mutations in patients who do not respond to EGFR-TKI and those resistant to EGFR-TKI.[31] Several studies have indicated that concomitant EGFR mutations affect the prognosis of patients with cancer. Chen et al[46] found that patients with concomitant mutations had significantly lower ORR (43.8 vs. 80.0%; P = 0.02) and PFS (P < 0.01) than those without concomitant mutations, after receiving EGFR-TKI treatment. In the BENEFIT phase 2 clinical trial, among patients with EFGR mutations, patients with single EGFR mutations had a better prognosis than patients with concomitant tumor suppressor gene mutations and concomitant oncogene mutations, with a median PFS duration of 13.2 (95% CI: 11.5–15.0) months, 9.3 (95% CI: 7.6–11.0) months, and 4.7 (95% CI: 1.9–9.3) months, respectively.[47] Liang et al[48] reported that patients with EGFR concomitant mutations had a poorer ORR than those with single EGFR mutations, both 19del (65.1 vs. 88.9%, P = 0.03) and L858R mutations (47.5 vs. 75.0%, P = 0.03). Wang et al[49] found that among patients with advanced EGFR ex20ins NSCLC, those with >1 concomitant mutation had a shorter PFS than those with ≤1 concomitant mutation (median PFS: 4.73 months vs. 10.74 months, P = 0.01), while no difference was found in OS.

Patient prognosis is also associated with different types of concomitant mutations. Blakely et al[31] showed that CDK4/6 alteration could promote tumor progression in patients receiving either first-generation or second-generation EGFR-TKIs (hazard ratio [HR]: 13.8, 95% CI: 5.2–36.9, P < 0.01) or osimertinib (HR: 10.3, 95% CI: 3.0–34.7, P < 0.01). Kim et al[50] found that for patients receiving first-/second-generation EGFR-TKIs, TP53 was a poor prognostic factor for PFS (P = 0.04) and OS (P < 0.01), while for patients with the T790M mutation receiving third-generation osimertinib, TP53 (PFS, P = 0.02; OS, P < 0.01), RB1 (PFS, P < 0.01; OS, P < 0.01), PTEN (PFS, P < 0.01), and MDM2 (PFS, P = 0.05; OS, P = 0.03) mutations all had adverse effects on prognosis. In summary, concomitant mutations affect cancer prognosis and correlate with the type of mutated genes.

Treatments for EGFR concomitant mutations Monotherapy

The EGFR-TKI therapy is a routine treatment regimen for patients with EGFR mutations. Zhuang et al[51] found that patients with advanced NSCLC with concomitant EGFR mutations (ALK, ROS1, KRAS, and BRAF mutations) receiving first-/second-generation EGFR-TKIs had a better prognosis than those receiving chemotherapy (median PFS: 10.8 vs. 5.2 months, P = 0.02). Furthermore, no significant difference was found between patients with concomitant EGFR mutations and patients with a single EGFR mutation receiving EGFR-TKIs (median PFS: 10.8 months vs. 9.6 months, P = 0.75, respectively). Hong et al[52] showed that the ORR for patients with advanced NSCLC with EGFR concomitant mutations receiving first-generation EGFR-TKIs was only 44%, with a median PFS of 6.2 (95% CI: 4.1–8.3) months and a median OS of 22.7 (95% CI: 18.0–22.3) months, which was much lower than that of patients without concomitant mutations (ORR: 77%, P = 0.01; median PFS: 18.8 months, P < 0.01, median OS: not reached, P < 0.01). Nomograms were also conducted to predict the prognosis of patients with EGFR mutations who received first-generation EGFR-TKIs, while concomitant mutations (including concomitant oncogenic driver gene mutations and concomitant tumor suppressor gene mutations) were found to be markers for poor prognosis in terms of both PFS and OS.[53] Besides, Jiang et al[54] reported that patients with EGFR concomitant mutations had a poorer prognosis than patients without concomitant mutations after receiving first-/second-generation EGFR-TKIs (patients with EGFR concomitant mutations vs. patients without concomitant mutations: median PFS: 7.4 months vs. 11.1 months, P = 0.04; median OS: 32.8 months vs. not reached, P = 0.01), as well as patients with T790M mutations after receiving osimertinib (patients with EGFR concomitant mutations vs. patients without concomitant mutations: median PFS: 3.6 months vs. 15.6 months, P < 0.01; median OS: 32.8 months vs. not reached, P = 0.03).

Similar inferior survival benefits were also found in several studies for specific concomitant EGFR mutations, including MET amplification, PTEN deletion, TP53 mutation, and PIK3CA mutations.[55–59] However, Labbé et al[60] found no significant survival difference associated with TP53 status in EGFR-mutant patients administered both first- and third-generation EGFR-TKIs, except patients with concomitant TP53 missense mutations (PFS: HR = 1.91, 95% CI: 1.01–3.60, P = 0.04). A phase Ib/II study also showed that patients with advanced EGFR mutations with concomitant TP53 mutations had PFS comparable to that of patients with wild-type TP53 (14.0 months vs. 15.4 months, P = 0.32) who received the novel second-generation EGFR-TKI mefatinib.[61] In addition, inconsistent results were also found in terms of B-cell lymphoma-2 interacting mediator of cell death (BIM) deletion polymorphism; Isobe et al[62] showed that EGFR-TKI use resulted in inferior survival for patients with concomitant BIM deletion (median PFS: 227 days vs. 533 days; P < 0.01) than for patients without BIM deletion, as did Zhao et al[63] (median PFS: 4.7 months vs. 11.0 months, P < 0.01; ORR: 25% vs. 66%, P < 0.01). However, both Ariyasu et al[64] and Liu et al[65] found that BIM mutations had no impact on PFS or OS, including in patients who received first-, second-, and third-generation EGFR-TKIs. Prior studies showed no significant difference between patients with EGFR/ALK co-alterations and patients with EGFR mutation receiving first-generation EGFR-TKIs (Zhao et al,[66] ORR: 63.2% vs. 62.1%, P = 0.93; median PFS: 10.3 months vs. 11.4 months, P = 0.87) (Lou et al,[67] ORR: 80.0% vs. 65.5%, P = 0.57; median PFS: 11.2 months vs. 13.2 months, P = 0.87). In addition, a retrospective study found that patients with concurrent EGFR mutations and HER2 amplification had better survival over patients with only EGFR mutations who received EGFR-TKIs (median OS: 846.0 days vs. 286.0 days, HR: 2.28, 95% CI: 1.29–4.05, P < 0.01).[68]

Thus, although EGFR-TKI monotherapy showed more survival benefits than chemotherapy for patients with concomitant mutations, in most cases, the prognosis was poorer than that of those without concomitant mutations, and the specific survival benefit should thus be based on the type of concomitant mutations. Considering the inconsistent results for several concomitant mutation studies, further explorations with larger sample sizes are needed.

Combination therapy

Combination therapy options mainly include EGFR-TKIs, other TKIs, chemotherapy, anti-angiogenic therapy, and immunotherapy, and prior studies have explored combination therapies with different TKIs. The MARQUEE study (NCT01244191) reported that for patients with advanced EGFR-mutant cancer who had previously received chemotherapy but not EGFR or MET inhibitors, the MET inhibitor tivantinib plus erlotinib increased ORR (60.7% vs. 43.4%) and PFS (median PFS: 13.0 months vs. 7.5 months, HR: 0.49, 95% CI: 0.31–0.77) when compared with erlotinib monotherapy.[69] Another real-world study found that patients with EGFR-mutant tumors with concomitant MET amplification after EGFR-TKI resistance achieved a superior ORR and PFS benefit when administrated EGFR-TKI plus crizotinib than when administered crizotinib monotherapy or chemotherapy alone (ORR: 48.6% vs. 40.0% vs. 18.2%; median PFS: 5.0 months vs. 2.3 months vs. 2.9 months, P = 0.01). However, no significance was found for OS benefit (10.0 months vs. 4.1 months vs. 8.5 months, P = 0.09).[70] In addition, the phase 1b study NCT02143466 explored the combination therapy of osimertinib and savolitinib for EGFR-mutant, MET-amplified patients who experienced disease progression after receiving EGFR-TKI treatment. Among patients previously receiving third-generation EGFR-TKIs, the ORR was 30% and median PFS was 5.4 (95% CI: 4.1–8.0) months. As for patients without previous third-generation EGFR-TKI treatment, the ORR was >60% and the median PFS was >9 months, while patients with the T790M mutation had a median PFS of 11.0 months (95% CI: 4.0 months to not reached).[71] In addition, a preclinical study also found that regarding the efficacy of combined TKI therapy, both osimertinib plus FGFR inhibitors (futibatinib or erdafitinib) and osimertinib plus crizotinib restored the sensitivity of patient-derived cell lines expressing the L858R mutantion plus FGFR3-TACC3 alterations and ex19del plus STRN-ALK alterations, respectively, and this finding was also validated in vivo.[72] Thus, the combination of two TKIs provides an alternative treatment strategy with corresponding targeted agents for concomitant mutations; however, specific indications and administrations should be further explored.

In addition, the ChiCTR1800016672 study explored the efficacy of the combination therapy of first-generation EGFR-TKIs and pemetrexed vs. first-generation EGFR-TKI monotherapy. Advanced treatment-naïve patients with EGFR-sensitive mutations but without T790M or ALK/MET/ROS1/RET fusion mutations were included. The combination therapy achieved a superior benefit over EGFR-TKI monotherapy (median PFS: 15.8 months vs. 5.0 months, P = 0.04) for patients with concomitant mutations (MLH1/V384D/TP53/KRAS/PIK3CA mutations), but no difference was found for patients without concomitant mutations (median PFS: 12.2 months vs. 11.1 months, P = 0.99).[73] Thus, EGFR-TKI plus chemotherapy might prolong PFS in patients with concomitant mutations such as MLH1/V384D/TP53/KRAS/PIK3CA mutations.

Furthermore, prior studies have explored the combination of anti-angiogenic therapy with EGFR-TKIs. For example, the ACTIVE study (CTONG1706/NCT02824458) explored the vascular endothelial growth factor receptor 2 TKI apatinib in combination with gefitinib for advanced treatment-naïve patients with concomitant EGFR mutations. The median PFS was 13.2 (95% CI: 7.0–19.4) months; therefore, apatinib plus gefitinib improved PFS when compared with gefitinib plus placebo for patients with TP53 exon 8 mutations (HR: 0.24, 95% CI: 0.06–0.91).[74,75] Moreover, the T790M mutation rate after disease progression was similar for patients with concomitant mutations who received apatinib plus gefitinib and gefitinib monotherapy (37.8% vs. 37.0%).[75] The RELAY study (NCT02411448) compared the efficacy of the combination therapy of ramucirumab plus erlotinib with erlotinib monotherapy and showed that for advanced treatment-naïve patients with EGFR and TP53 mutations, combination therapy improved PFS for both patients with ex19del (PFS: 18.0 months vs. 9.9 months, HR: 0.50, 95% CI: 0.29–0.85) and those with L858R mutations (PFS: 14.7 months vs. 10.8 months, HR: 0.56, 95% CI: 0.34–0.95). However, for EGFR mutant patients with TP53 wild-type tumors, neither the ex19del (combination therapy vs. monotherapy: PFS: 20.6 months vs. 19.4 months, HR: 1.00, 95% CI: 0.62–1.61) nor L858R (combination therapy vs. monotherapy: PFS: 20.8 months vs. 13.8 months, HR: 0.60, 95% CI: 0.35–1.02) cohorts showed significant differences.[76] Moreover, Zhang et al[53] found that first-generation EGFR-TKI plus bevacizumab prolonged survival for patients with EGFR-mutant advanced NSCLC harboring tumor suppressor gene mutations when compared with EGFR-TKI monotherapy (median PFS: 14.5 months vs. 8.0 months, P < 0.01), while no significance was found for patients harboring oncogenic driver gene mutations. The predictive nomogram model also showed that the presence of concomitant mutations did not affect the prognosis of patients receiving anti-angiogenic therapy plus EGFR-TKIs in terms of both PFS and OS.[53] Therefore, the combination of anti-angiogenic therapy and EGFR-TKIs is also an optional treatment strategy for patients with concomitant mutations.

In additio