Therapeutic Impact of Third-generation Tyrosine Kinase Inhibitors on Lung Squamous Cell Carcinoma with Epidermal Growth Factor Receptor Mutations: A Case Description and Systematic Literature Review

Abstract

Epidermal growth factor receptor (EGFR) mutations are uncommon in lung squamous cell carcinoma (SQC). However, the role of third-generation EGFR-tyrosine kinase inhibitors (TKIs) remains unclear. The present study provides a case description and a systematic literature review of 34 eligible cases. Third-generation EGFR-TKIs demonstrate superior initial responsiveness compared to earlier-generation TKIs. The efficacy of TKIs was consistent across SQC forms (including de novo or adenocarcinoma-transformed cases) and EGFR mutation types. The findings indicate that, compared with early-generation TKIs, third-generation TKIs are effective against de novo SQCs and should be considered a plausible treatment option for transformed SQCs.

Introduction

Third-generation tyrosine kinase inhibitors (TKIs) targeting the epidermal growth factor receptor (EGFR) are the standard therapeutic options for patients with non-small-cell lung cancer (NSCLC) harboring EGFR mutations. Sensitive EGFR mutations, including a genetic deletion in exon 19 (Ex19del) and the L858R point mutation in exon 21 (L858R), are prevalent in lung adenocarcinomas (ADCs), particularly in the Eastern Asian population.

A large-scale clinical trial revealed that osimertinib, a third-generation EGFR-TKI, demonstrated high sensitivity to ADCs with sensitive EGFR mutations and provided clinical benefits in improving the progression-free survival (PFS) and overall survival (OS) in patients with untreated ADCs compared with first- or second-generation TKIs (1). Conversely, genetic mutations in EGFR are relatively rare in lung squamous cell carcinomas (SQCs), with an incidence of 6.9-14.7% in the Asian population (2-4).

Previous research has demonstrated that first- and second-generation TKIs exhibit reduced sensitivity to SQCs (or non-ADC-NSCLCs) harboring sensitive EGFR mutations in cohort studies (3,5-9). A subsequent pooled analysis of reported cases further revealed that gefitinib, a first-generation EGFR-TKI, demonstrated an inferior objective response rate (ORR; 35% vs. 69%), disease control rate (DCR; 80% vs. 94-95%), and PFS (3.1 vs. 9.8 months) in non-ADC-NSCLCs compared to ADCs, both with sensitive EGFR mutations (10). However, patients with SQCs with EGFR mutations who underwent treatment with first- or second-generation TKIs exhibited a superior prognosis to those with SQCs with EGFR mutations who did not undergo TKI therapy (3,5-7) or those with EGFR-wild-type SQCs who underwent TKI therapy (3,5,11). These findings suggest that a subset of patients with SQCs and EGFR mutations may benefit from TKI therapy. However, studies evaluating the efficacy of third-generation EGFR-TKIs in SQCs with EGFR mutations are scarce, and their therapeutic impact remains debatable.

A small cohort study of 7 patients with SQCs with EGFR mutations demonstrated an ORR of 57.1%, a DCR of 85.7%, and a PFS of 0-9.6 months with osimertinib (12). In addition, a series of case studies and reviews demonstrated a wide variety of responses to osimertinib in SQCs with EGFR mutations (13,14). Notably, SQCs with EGFR mutations manifest in two distinct forms in studies involving third-generation TKIs: cases with de novo mutations present at the diagnosis (de novo SQCs), and cases arising from the transformation of ADCs acquired during TKI therapy (transformed SQCs).

To our knowledge, the present study is the first to conduct a systematic case review assessing the efficacy of third-generation EGFR-TKIs across different SQC forms, EGFR mutation types, and the presence or absence of a T790M co-mutation.

Case Report

An 84-year-old Japanese man was admitted to our hospital with dyspnea on exertion, demonstrating a moderate amount of right-sided pleural effusion. The patient's vital signs were unremarkable.

Laboratory examinations revealed elevated serum levels of carcinoembryonic antigen (6.8 ng/mL), sialyl Lewis X (320 U/mL), squamous cell carcinoma antigen (18.1 ng/mL), and cytokeratin 19 fragment (46 ng/mL). Computed tomography (CT) revealed a lung tumor in the right lower lobe with direct invasion of the diaphragm and liver; pleural nodules; a moderate amount of right-sided pleural effusion; and hilar, mediastinal, and supraclavicular lymph node enlargement (Fig. 1A, B). A histological examination of a transepidermal needle biopsy of the liver tumor revealed nonkeratinized SQC on hematoxylin and eosin staining (Fig. 2A). Immunohistochemical examinations revealed positive staining for p40 and negative staining for TTF-1 or Napsin A (Fig. 2B-D), indicating SQC. Next-generation sequencing using Oncomine multi-CDx (Thermo Fisher Scientific, Waltham, USA) revealed an EGFR mutation at L858R in the liver biopsy specimen. The tissue proportion score for programmed cell death ligand-1 was 5% using the anti-programmed cell death ligand-1 antibody clone 22C3. The patient was thus ultimately diagnosed with lung SQC at T4N3M1a, clinical stage IVA (9th edition TNM classification for lung cancer) (15).

Figure 1.

Radiological findings. Chest computed tomography revealed a lung tumor in the right lower lobe with direct invasion of the diaphragm and liver (black arrows) and right-sided pleural effusion before osimertinib treatment (A and B). The tumor demonstrated regression one month after the initiation of osimertinib treatment (black arrows) (C and D). After three months of osimertinib treatment, the tumor shrank further (black arrows) (E and F). Concurrently, multiple metastases are observed in the vertebrae (white arrows) (G and H).

Figure 2.

Histological examinations. A histological evaluation of a transepidermal needle biopsy of the liver tumor revealed non-keratinized squamous cell carcinoma, as indicated by Hematoxylin and Eosin staining (A). Immunohistochemical examinations revealed positive staining for p40 (B) and negative staining for TTF-1 (C) and Napsin A (D). Original magnification ×400.

The patient underwent thoracic drainage followed by pleurodesis with a talc slurry. The Eastern Cooperative Oncology Group (ECOG) performance status score was 2. Given the patient's health status and treatment preferences, osimertinib monotherapy (80 mg once daily) was initiated as the first-line treatment regimen. Chest CT revealed substantial regression of the lung-hepatic mass, pleural nodules, and lymph nodes after four weeks of osimertinib treatment (Fig. 1C, D), indicating a partial response based on the Response Evaluation Criteria in Solid Tumors version 1.1. No adverse effects were reported, and the performance status improved to 1. Subsequent CT revealed further shrinkage of the primary tumor in the lung-hepatic mass (Fig. 1E, F). However, multiple bone metastases developed in the vertebrae (Fig. 1G, H), indicating disease progression 12 weeks after the initiation of osimertinib therapy.

The patient and his family declined to pursue subsequent treatment with cytotoxic agents. The patient continued to receive osimertinib beyond disease progression to limit the tumor burden in the lungs and liver.

Systematic Case Reviews

Search strategy

A systematic literature review was conducted using PubMed to investigate the use of third-generation EGFR-TKIs in the treatment of SQCs with EGFR mutations. A flow diagram of the case study is shown (Fig. 3). The initial search was performed using the following search terms: [“lung” (MeSH Terms) OR “lung” (Text Words)] AND [“squamous” (Text Words) OR “non-adenocarcinoma” (Text Words)] AND [“carcinoma” (MeSH Terms) OR “carcinoma” (Text Words) OR “cancer” (Text Words)] AND [“ErbB receptors” (MeSH Terms) OR “epidermal growth factor receptors” (Text Words) OR “epidermal growth factor receptor” (Text Words) OR “EGFR” (Text Words)] AND [“tyrosine kinase inhibitors” (MeSH Terms) OR “tyrosine kinase inhibitors” (Text Words) OR “tyrosine kinase inhibitors” (Pharmacological Action) OR “tyrosine kinase inhibitor” (Text Words) OR “TKI” (Text Words) OR “TKIs” (Text Words)]. The search period was restricted to the period from 2015, when third-generation EGFR-TKIs were developed, to August 14, 2025. Patients with de novo lung adenosquamous cell carcinoma were also excluded. However, cases with an intermediate status through transformation from ADCs to SQCs were included.

Figure 3.

A flow diagram of case inclusion.

Results

The database search identified 573 studies. Of these, 46 involved case reports or series including SQCs with EGFR mutations that had been treated with TKIs of any type. Full-text screening confirmed 30 eligible cases of third-generation TKIs in 27 studies. The remaining three cases were incorporated into the analysis after identification through a search using Ichu-shi, a Japanese search engine.

The characteristics of the 33 registered patients and the present case are summarized (Supplementary material). Among them, 32 were treated with osimertinib, 1 with aumolertinib, and 1 with almonertinib. The response to third-generation TKIs was analyzed in 29 cases, with the exclusion of 5 cases due to the concurrent use of other antitumor agents or radiotherapy. The majority of the studies were conducted in East Asia (31 cases; 91.1%), and the study population included individuals age 39-85 (median age: 63.5) years old, women (13 cases; 38.2%), and never-smokers (17 cases; 50.0%) (Table 1). Nine patients had not received prior treatment, whereas the remaining 25 had received 1-6 treatment lines before utilizing third-generation TKIs (Supplementary material). Among these, early-generation TKIs were prescribed to 22 patients. Five patients experienced tumor recurrence after radical surgery or concurrent chemoradiotherapy. Third-generation TKI treatment demonstrated an ORR of 65.5% and a DCR of 89.6% (Table 1). The absolute PFS was analyzed in 15 patients, yielding a median duration of 3.0 (range: 0-24.0) months. In the other 13 cases, the TKI treatment was ongoing without progression at the time of the report (censored patients), demonstrating a median PFS of 4.0 (range: 0.75-48.0) months. A combined analysis revealed a median PFS of 3.5 months in 28 patients (Table 1). However, the available data regarding the OS were limited, resulting in an incomplete evaluation.

Table 1.

Case Summary of Squamous Cell Carcinomas with EGFR Mutation Treated with Third-generation Tyrosine Kinase Inhibitors.

	n	Age (median) (years)	Females	Non-smokers	CR/PR/SD/PD*	ORR*	DCR*	PFS$
total	34	39-85 (63.5)	13 (38.2%)	17 (50.0%)	1/18/7/3	65.5%	89.6%	3.5 months
de novo SQCs	18	39-85 (64.0)	8 (44.4%)	10 (55.5%)	1/10/3/3	64.7%	82.3%	3.0 months
Transformed SQCs	16	40-80 (62.5)	5 (31.2%)	7 (43.7%)	0/8/4/0	66.6%	100%	3.0 months
Ex19del	22	39-83 (64.0)	7 (31.8%)	10 (45.4%)	1/12/4/1	72.2%	94.4%	5.0 months
L858R	10	47-85 (66.0)	5 (50.0%)	6 (60.0%)	0/6/2/2	60.0%	80.0%	3.0 months
others	2	54-61 (57.5)	1 (50.0%)	1 (50.0%)	0/0/1/0	0%	100%	3.0 months
T790M (+)	23	39-80 (60.0)	6 (26.0%)	9 (39.1%)	0/13/6/0	68.4%	100%	6.0 months
T790M (-)	11	54-85 (70.0)	8 (72.7%)	8 (72.7%)	1/5/1/3	60.0%	70.0%	2.0 months

*: Analyzed in 29 cases.

^$: Analyzed in 28 cases at the time of the report.

CR: complete response, DCR: disease control rate, EGFR: epidermal growth factor receptor, ORR: objective response rate, PD: progressive disease, PFS: progression-free survival, PR: partial response, SD: stable disease, SQCs: squamous cell carcinomas

Third-generation TKIs were prescribed to patients with de novo SQCs (n=18) or transformed SQCs (n=16) (Table 1). The two groups exhibited comparable patterns of EGFR mutation. The de novo SQC group had a slightly higher proportion of women (44.4% vs. 31.2%) and never-smokers (55.5% vs. 43.7%) than the transformed SQC group. A comparative analysis yielded analogous results for the ORR (64.7% vs. 66.6%) and PFS (3.0 vs. 3.0 months) between the de novo and transformed SQC groups. An inferior DCR was observed in de novo SQCs compared with transformed SQCs (82.3% vs. 100%) due to 3 progressive disease cases in the former. The EGFR gene mutations were Ex19del (22 cases; 64.7%), L858R (10 cases; 29.4%), and uncommon (2 cases; 5.8%), indicating a relatively high prevalence of Ex19del in this population (Table 1). The Ex19del group had a lower proportion of women (31.8% vs. 50.0%) and non-smokers (45.4% vs. 60.0%) than the L858R group. A marginally higher ORR (72.2% vs. 60.0%) and DCR (94.4% vs. 80.0%), along with a slightly longer median PFS (5.0 vs. 3.0 months), were observed in the Ex19del group than in the L858R group. The co-mutation of T790M was detected in 8 of 18 patients (2: de novo, 6: acquired forms) with de novo SQCs and in 15 of 16 patients (all acquired form) with transformed SQCs. The T790M-positive group had a lower proportion of women (26.0% vs. 72.7%) and non-smokers (39.1% vs. 72.7%) than the T790M-negative group. The presence (vs. absence) of the T790M mutation did not affect the ORR (68.4% vs. 60.0%), DCR (100% vs. 70.0%), or median PFS (6.0 vs. 2.0 months).

Finally, intergroup comparisons were performed to examine the variables influencing treatment efficacy; the results are presented (Table 2). Clinical parameters were compared between groups with (complete or partial response) and without (stable or progressive disease) treatment response using Fisher's exact test for nominal variables or the Mann-Whitney U-test for continuous variables. The analysis was performed using EZR (Jichi Medical University Saitama Medical Center, Saitama, Japan) (16). Consequently, the treatment efficacy of third-generation TKIs for SQCs with EGFR mutations was not influenced by any variables, including the age, sex, smoking history, TKI treatment history, SQC form, type of EGFR mutations, or presence of a T790M co-mutation (Table 2).

Table 2.

Comparison of Clinical Parameters between Different Therapeutic Responses.

	All patients (n=29)	CR/PR (n=19)	SD/PD (n=10)	p value
Age (years) [median (25%, 75%)]	65.0 (57.0, 72.0)	65.0 (58.5, 70.5)	64.5 (55.7, 71.5)	0.748
Sex (Male/female)	19/10	11/8	8/2	0.414
Smoking history (Yes/No)	16/13	10/9	6/4	1.00
Prior TKI treatment (No/Yes)	11/18	8/11	3/7	0.694
Form of SQCs (de novo/transformed)	17/12	11/8	6/4	1.00
Types of mutation (Ex19del/L858R)a	18/10	13/6	5/4	0.677
T790M mutation (No/Yes)	10/19	6/13	4/6	0.698

^a: Analyzed in 28 cases.

CR: complete response, PD: progressive disease, PR: partial response, SD: stable disease, SQCs: squamous cellcarcinoma, TKI: tyrosine kinase inhibitor

Discussion

The present study constitutes an exhaustive case review of the literature, with the objective of summarizing the current knowledge on the treatment of SQCs with EGFR mutations using third-generation TKIs, encompassing 34 eligible cases. The study population exhibited a prevalence of women (38.2%) and non-smokers (50.0%), analogous with the findings of previous studies (8,11,17). These studies demonstrated that, in comparison with EGFR-wild-type SQCs, SQCs with EGFR mutations exhibited a higher prevalence in women (40.7-76.9% vs. 4.4-52.5%) and non-smokers (53.7-83.3% vs. 11.1-30.0%). However, inconsistent results have been reported (7).

The ORR of 65.5% and DCR of 89.6% obtained in our study were comparable to those of a cohort study using third-generation TKIs in patients with similar characteristics (12). The initial responses, i.e. ORR and DCR, in the present study were more favorable than the findings from studies employing first- or second-generation TKIs for SQCs with EGFR mutations (31.6-44.7% and 72.0-81.8%, respectively) (3,5-8), despite the indirect nature of the comparison among different populations and study designs. An analogous trend has been reported in ADCs with EGFR mutations, demonstrating a superior ORR with third-generation TKIs (70%) over early-generation TKIs (58%) (18). Furthermore, the ORR in the present study was marginally lower than but close to that observed in studies involving ADCs with EGFR mutations treated with osimertinib (1,18). Nevertheless, it is reasonable to hypothesize the presence of selection bias, with a tendency for increased case reporting in studies demonstrating a favorable response.

Conversely, the median PFS of 3.5 months in our review exhibited equivalence with that in the studies employing early-generation TKIs for SQCs with EGFR mutations (3.0-8.0 months) (3,5-8,10) and was considerably shorter than that observed in ADCs with third-generation TKIs (18.9-23.4 months) (1,18). These findings suggest that third-generation TKIs may demonstrate superior initial responses while maintaining a short duration of response in patients with SQC harboring EGFR mutations when compared to early-generation TKIs. However, the evaluation of PFS in the case reviews was premature and may have been underestimated because the PFS analysis included many censored patients (13/28 cases). Further research is warranted to clarify the long-standing benefits of third-generation TKI treatment for SQCs.

The secondary objective of the present study was to examine the efficacy of TKI treatment in de novo and transformed SQCs separately. This approach was motivated by the hypothesis that these two types of SQCs may exhibit unique characteristics. For instance, transformed small-cell carcinomas from ADCs with EGFR mutations exhibit partially shared yet distinct gene expression and molecular characteristics from those of the original de novo small-cell carcinoma, necessitating a different treatment strategy (19,20). Conversely, the literature on the pathophysiological similarities or differences between distinct forms of SQCs is limited, partly because of the low incidence of SQC transformation. The maintenance of the initial forms of EGFR mutations in the transformed SQCs, as observed in the present review, supports the lineage evolution theory rather than the selection pressure for the mechanisms underlying the histological transformation from ADCs to SQCs during TKI treatment. In this scenario, transformed SQCs may exhibit unique genomic profiles from de novo SQCs and potentially manifest a distinct nature in their response to treatment. Consequently, a comparative analysis between the two forms was conducted in the present study, thereby demonstrating, for the first time, that third-generation TKIs exhibit similar efficacy in both the de novo and transformed forms of SQCs with EGFR mutations (Tables 1, 22). These findings indicate that compared with early-generation TKIs, the use of third-generation TKIs is more feasible for de novo SQCs.

Furthermore, third-generation TKIs should be considered a plausible treatment option for transformed SQCs where early-generation TKIs are not applicable because of the frequent existence of a T790M co-mutation. However, it is imperative to acknowledge that the majority of cases examined in this study exhibited an acquired T790M mutation in conjunction with SQC transformation during the failure of TKI therapy, suggesting that transformed SQCs still depend on the EGFR pathway. In contrast, several investigators have identified genomic alterations in the phosphoinositide-3-kinase/mammalian target of rapamycin pathway or amplification of driver oncogenes other than EGFR in post-transformed SQCs (21-23), indicating a mechanism of TKI resistance not dependent on the EGFR pathway. In these cases, the efficacy of third-generation TKIs is expected to be low. These findings indicate the necessity for a comprehensive search for genomic alterations in transformed SQCs to determine the indications for third-generation TKIs.

Another factor that influences the efficacy of EGFR-TKIs in ADCs is the type of EGFR mutation. In a clinical trial, patients with ADC possessing the Ex19del mutation exhibited an ORR or DCR comparable to those with the L858R mutation (1). Other studies have demonstrated that patients with ADC and the Ex19del mutation exhibit a higher ORR than those with the L858R mutation when treated with first- (24,25) or third-generation (26) TKIs. The findings of the present study indicate that SQCs with the Ex19del mutation exhibited a modestly superior, yet small impact on the ORR, DCR, and median PFS compared to SQCs with the L858R mutation (Table 1). Furthermore, the type of EGFR mutation was not a variable that predicted the treatment efficacy (Table 2). In the present case review, the T790M co-mutation did not affect responsiveness to third-generation TKIs. This observation is analogous to findings in ADC cases, which demonstrated an ORR of 80% with first-line osimertinib treatment (1) and 71% with late-line treatment in patients harboring the T790M co-mutation (27).

Several limitations associated with the present study warrant mention. First, inherent selection bias in reported cases when estimating drug efficacy must be acknowledged. Second, the case reviews included a heterogeneous population with regard to the clinical characteristics and evaluation protocols. This may have affected the evaluation of the best response or PFS. Third, the PFS analysis was premature during the case review owing to the inclusion of numerous censored patients. Finally, the sample size was inadequate to conduct statistical analyses for intragroup comparisons. Therefore, the findings of this study should be cautiously interpreted.

In summary, the present study demonstrated that third-generation TKIs may improve the initial response compared to early-generation TKIs, while providing an immature evaluation of long-term effects. The present study suggests that the use of third-generation TKIs is more feasible than early-generation TKIs for de novo SQCs and should be considered a plausible treatment option for transformed SQCs. This review did not identify any relevant variables for predicting the effectiveness of third-generation TKIs in SQCs with EGFR mutations, including the form of SQCs (de novo or ADC-transformed). Further research is warranted to better understand the role of third-generation TKIs in SQCs harboring EGFR mutations.

The authors state that they have no Conflict of Interest (COI).

Supplementary Material

Supplemental Table 1.

Clinical characteristics of squamous cell carcinomas with EGFR mutation treated with third-generation tyrosine kinase inhibitors