Clinicopathologic features and therapeutic outcomes in EGFR/ALK-aberrations NSCLC with histologic transformation Authors: Kaibo Ding, Xinyue Li, Hui Li, Jing Qin, Dujiang Liu, Yanjun Xu Published: 2025 ABSTRACT AbstractBackgroundThe aim of this study is to delineate clinicopathologic features, therapeutic outcomes, and tumor microenvironment (TME) remodeling in non-small cell lung cancer (NSCLC) patients with epidermal growth factor receptor/anaplastic lymphoma kinase (EGFR/ALK) aberrations who experience histologic transformation (HT), with emphasis on post-transformation treatment efficacy.MethodsThis single-center retrospective cohort study included 6,230 patients with advanced NSCLC treated between December 2017 and January 2024. Patients who experienced histologic transformation after progression were identified through high-quality tumor biopsies or cytological samples. Treatment regimens for post-HT, including chemotherapy combined with tyrosine kinase inhibitors (CTK), chemotherapy combined with anti-angiogenic/immunotherapy (CAI), and palliative care (PC), were analyzed. Multiplex immunofluorescence was used to profile the tumor microenvironment in a subset of patients.ResultsOf the 6,230 patients, 84 (1.35%) experienced HT, with 73.81% transforming to small cell lung cancer (SCLC). The median time to HT was 22.27 months. Patients receiving CTK exhibited significantly longer median overall survival (OS1) after HT (OS1: 15.40 months) compared to CAI (11.23 months) and PC (4.37 months; P = 0.003). While CTK improved OS1, progression-free survival (PFS1) did not differ significantly across groups. In second-line treatment, incorporating tyrosine kinase inhibitors (TKIs) significantly prolonged PFS2 (PFS2; median: 6.70 months, compared with 2.87 months for regimens not incorporating TKIs; P = 0.016). Uncommon EGFR mutants exhibited paradoxical tumor microenvironment (TME) features, with elevated CD8+ infiltration (128.5 vs. 64.2 cells/mm²; p = 0.028) accompanied by Treg dominance (Foxp3+CD25+/CD8+ ratio: 0.41 vs. 0.19; p = 0.039). Immunotherapy-containing regimens improved PFS1 in uncommon mutants (3.10 vs. 1.83 months; P = 0.007); however, overall survival remained comparable. Conclusions: Histologic transformation poses a significant challenge in advanced NSCLC management. Chemotherapy combined with TKIs appears to offer the best survival benefit for post-HT, while second-line inclusion of TKIs improves PFS. Personalized strategies accounting for histologic subtype, molecular alterations, and immune contexture are essential for optimizing outcomes in this clinically distinct population. Further large-scale studies are warranted to refine therapeutic approaches.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12967-025-07540-w. Introduction Non-small cell lung cancer (NSCLC) remains a leading cause of cancer-related mortality worldwide, with approximately 50% of patients presenting with advanced or metastatic disease at diagnosis [1, 2]. The advent of molecularly targeted therapies, particularly epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) and anaplastic lymphoma kinase inhibitors (ALK-TKIs), has revolutionized treatment paradigms for oncogene-driven NSCLC, significantly improving progression-free survival (PFS) and overall survival (OS) [3–5]. However, the emergence of acquired resistance mechanisms, including histologic transformation (HT), poses a critical clinical challenge. HT, which is characterized by lineage plasticity leading to phenotypic shifts such as small cell lung cancer (SCLC) or other NSCLC subtypes, has been increasingly recognized as a resistance mechanism in 3–15% of EGFR-mutant NSCLC cases following TKI therapy [6–9]. This phenomenon not only alters tumor biology but also complicates therapeutic decision-making due to divergent treatment responses compared to the original histology [10, 11]. Recent studies have highlighted the molecular underpinnings of HT, particularly the enrichment of Tumor Protein 53 (TP53) and Retinoblastoma 1 (RB1) co-mutations in SCLC-transformed cases, which may drive neuroendocrine differentiation and therapeutic recalcitrance [11–13]. Despite these advances, consensus on optimal post-HT management remains elusive. Current guidelines lack specific recommendations for HT, and clinical evidence is largely derived from small retrospective cohorts or case series [10, 14, 15]. While etoposide–platinum chemotherapy remains the mainstay for SCLC-transformed cases, emerging strategies combining chemotherapy with targeted agents (e.g., TKIs) or immunotherapies have shown variable efficacy in limited studies [16–19]. Furthermore, the clinical behavior and therapeutic vulnerabilities of rare EGFR-mutant subtypes undergoing HT are poorly characterized, underscoring the need for large-scale analyses to inform precision approaches. In this single-center retrospective cohort study, we analyzed 6,230 advanced NSCLC patients to delineate the incidence, clinicopathologic features, and therapeutic outcomes of HT, with a focus on post-transformation treatment efficacy. We specifically evaluated the impact of chemotherapy combined with TKIs (CTK) versus chemotherapy with anti-angiogenic/immunotherapy (CAI) on survival outcomes, leveraging one of the largest HT cohorts reported to date. Additionally, we explored the tumor microenvironment (TME) in a subset of 13 patients through multiplex immunohistochemistry, providing preliminary insights into immune cell infiltration patterns for post-HT. Our findings aim to address critical knowledge gaps in HT management and refine therapeutic strategies for this clinically distinct population. Materials and methods Study design and participants This single-center, retrospective, hospital-based cohort study enrolled patients with advanced gene-mutated lung cancer treated at our institution between December 2017 and January 2024. Inclusion criteria were as follows: (1) diagnosis of advanced or recurrent stage IV NSCLC per the 9th edition of the TNM classification [20], with gene mutations confirmed by polymerase chain reaction (PCR) or next-generation sequencing (NGS); (2) HT confirmed via high-quality tumor biopsies or well-preserved cytological samples during treatment, followed by receipt of at least two post-HT therapy cycles at our center. Exclusion criteria included cases transitioning from mixed histology to a single histology (e.g., adenosquamous carcinoma to squamous cell carcinoma or combined SCLC to pure SCLC) [21]. Post-HT pathologic subtypes were categorized into two groups: transformation to SCLC and transformation to other NSCLC subtypes (including squamous cell carcinoma [SCC], pulmonary sarcomatoid carcinoma [PSC], and neuroendocrine carcinoma not otherwise specified [NE-NOS]), based on the 2015 WHO classification of thoracic tumors [22]. Data collection included demographic information, baseline histology, immunohistochemistry, molecular pathology, transformed histopathology, treatment regimens, adverse events, and survival outcomes. Gene profiling at the initial NSCLC diagnosis was performed using various methods, such as allele-specific PCR and NGS. The authors take full responsibility for all aspects of the work, ensuring that any issues related to the accuracy or integrity of the study are appropriately investigated and resolved. In accordance with the Declaration of Helsinki, and due to the retrospective nature of the study, the requirement for written informed consent was waived by the Ethics Committee. The study encompassed patients treated between December 2017 and Jan-uary 2024. Definitions and assessments HT: Defined as a change in the original histologic characteristics of advanced lung cancer to another histologic type following disease progression or treatment. Time to HT: Defined as the interval from the initial pathologic diagnosis of advanced lung cancer to the identification of another histologic type through additional biopsies. HT Incidence: Calculated as the ratio of the number of transformation cases to the number of cases with histologic confirmation via re-biopsy after disease progression. Progression-Free Survival (PFS1): Defined as the time from the initiation of first-line treatment after HT to disease progression or death. Overall Survival (OS): Defined as the time from the diagnosis of advanced NSCLC to death or the last follow-up assessment. Overall Survival After HT (OS1): Defined as the time from the diagnosis of HT to death or the last follow-up assessment. Progression-Free Survival After Second-Line Treatment (PFS2): Defined as the time from the initiation of second-line treatment after HT to disease progression or death. Treatment and response evaluation criteria For patients with advanced NSCLC transforming to SCLC, treatment information after transformation was collected. Tumor responses were evaluated every 6–8 weeks according to the Response Evaluation Criteria in Solid Tumors (RECIST) guidelines (version 1.1). Objective tumor responses were categorized as complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD). The disease control rate (DCR) was defined as the sum of the objective response rate and stabilization rate (CR + PR + SD). The choice of second-line treatment was not protocol-defined but based on the treating physician’s comprehensive assessment, which considered factors such as the patient’s performance status, prior treatment history and response, and the specific histologic and molecular characteristics of the transformed tumor. Multiplex immunofluorescence [23] Tissue processing and multiplex immunofluorescence staining Formalin-fixed paraffin-embedded (FFPE) tumor specimens from 13 patients (7 with uncommon EGFR mutations and 6 with classical EGFR-sensitive mutations) were sectioned into 4 μm slices using a rotary microtome (Leica HistoCore BIOCUT). Sections underwent sequential dewaxing in xylene I/II (12 min each) and rehydration through graded ethanol (100% to 85%, 6 min each). Antigen retrieval was performed in EDTA buffer (pH 8.0) using microwave heating (medium power: 8 min on, 8 min off; low power: 7 min), followed by cooling and PBS washing (3 × 5 min). Endogenous peroxidase activity was blocked with 3% H₂O₂ (25 min, RT). Tissue boundaries were outlined with a hydrophobic barrier pen (Gene tech GT1001) to prevent antibody dispersion. Multiplex fluorescence labeling Sections were blocked with 3% BSA (30 min, RT) and incubated with primary antibodies at 4 °C overnight. After PBS washing (3 × 5 min), HRP-conjugated secondary antibodies (Haoke Biotechnology, HKI0029) were applied (50 min, RT). Tyramide signal amplification (TSA) was performed using Flare520 (green), Flare570 (red), or Flare620 (orange) reagent (Haoke Biotechnology, 3–5 min). Sequential rounds of antibody stripping (microwave heating in EDTA buffer), blocking, and labeling were conducted for multiplex detection. Nuclei were counterstained with DAPI (Haoke Biotechnology, HKI0005), and slides were mounted with anti-fade medium (Haoke Biotechnology, HKI0007-1). Image acquisition and quantitative analysis Fluorescence images were captured using a Nikon Eclipse C1 microscope with predefined emission/excitation wavelengths (DAPI: 350/420 nm; Flare520: 490/520 nm; Flare570: 550/570 nm; Flare620: 590/620 nm). Whole-slide scanning was performed on a 3D HISTECH Pannoramic SCAN system. Quantitative analysis utilized Visiopharm AI-based software (v2023.1) for automated cell segmentation and marker quantification. Adaptive thresholds were applied to exclude background noise, and batch consistency was ensured by uniform analysis parameters across all samples. Statistical analysis Survival analyses were conducted using the Kaplan–Meier method, and differences in survival curves were assessed using the log-rank test. Hazard ratios (HRs) and corresponding 95% confidence intervals (CIs) were estimated using the Cox proportional hazards model. Univariate analysis was performed to screen variables, and those with statistically significant p-values were included in the subsequent multivariate Cox regression model. All p-values were two-sided, with statistical significance set at p < 0.05. For multiple subgroup analyses and comparisons, Bonferroni correction was performed to control the type I error, with statistical significance set at adjusted P < 0.05. Statistical analyses were performed using SPSS software (version 25; IBM Corporation, Armonk, NY, USA). Results Demographic and clinical characteristics before transformation A total of 6,230 patients with advanced NSCLC were enrolled, among whom 84 (1.35%) experienced histologic transformation (HT). The types of transformation were categorized into two groups: transformation from NSCLC to SCLC (n = 62; 73.81%) and transformation to another NSCLC subtype (n = 22; 26.19%). The latter group included squamous cell carcinoma (n = 11; 50.00%), pulmonary sarcomatoid carcinoma (n = 2; 9.09%), and neuroendocrine carcinoma not otherwise specified (n = 9; 40.91%). Detailed information on HT is provided in Supplementary Table 1. Table 1 summarizes the characteristics of patients with HT at the time of advanced lung cancer diagnosis. The median age was 62.0 years, with 51 male patients (60.7%). Most patients (n = 54; 64.3%) had a performance status (PS) of 0, and the majority were non-smokers (n = 57; 67.9%). All patients were diagnosed with stage IV NSCLC according to the 9th edition of the AJCC Cancer Staging Manual. Initially, 83 patients (98.81%) were diagnosed with adenocarcinoma, while only one patient was diagnosed with squamous cell carcinoma. EGFR mutations were present in 78 patients (92.86%), including exon 19 deletions (n = 42; 53.85%), L858R mutations (n = 25; 32.05%), and rare EGFR mutations (n = 11; 14.10%). Additionally, six patients had EML4-ALK fusion mutations. Among the 50 patients who underwent next-generation sequencing (NGS), 22 had TP53 co-mutations, and 20 of the TP53-mutated patients also harbored RB1 mutations. Table 1Characteristics of histologically transformed patients at the time of diagnosis of advanced NSCLCCharacteristicsTotal(n = 84)Patients with HT to SCLC(n = 62)Patients with HT to another NSCLC subtype(n = 22)P-ValueAgeMedian (range)62(33–83)62 (36–79)61 (33–83)0.452Sex, n0.752 Male513714 Female33258Smoking history, n0.351 Never smoking574017 Ever smoking27225Performance status, n0.414 0544113 130219Histology, n1.000 Adenocarcinoma836122 Squamous110Brain metastasis, n0.144 Yes1165 No735617Type of mutation, n0.568EGFR 19 deletion423012 L858R25214 Rare mutations1183 EML4-ALK fusion633Abbreviations: HT, histologic transformation; SCLC, small cell lung cancer; NSCLC, non- small cell lung cancer; EGFR, epidermal growth factor receptor; ALK, anaplastic lymphoma kinase. Among 50 patients who underwent next-generation sequencing (NGS), 22 had TP53 co-mutations, and 20 of the TP53-mutated patients also harbored RB1 mutations. P-value: chi-squared test for categorical variables Characteristics of histologically transformed patients at the time of diagnosis of advanced NSCLC Age Median (range) Abbreviations: HT, histologic transformation; SCLC, small cell lung cancer; NSCLC, non- small cell lung cancer; EGFR, epidermal growth factor receptor; ALK, anaplastic lymphoma kinase. Among 50 patients who underwent next-generation sequencing (NGS), 22 had TP53 co-mutations, and 20 of the TP53-mutated patients also harbored RB1 mutations. P-value: chi-squared test for categorical variables Treatment and its efficacy in patients before HT The treatment details before HT are presented in Supplementary Table 2. The median number of treatment lines before HT was 2 (IQR, 1–3.25), and the median time to HT was 22.27 months (95% CI: 18.28–26.28 months). In the first-line setting before HT, 33 patients (39.29%) received first- or second-generation EGFR-TKIs, while 30 patients (35.71%) received third-generation EGFR-TKIs. Five patients were treated with ALK-TKIs (Crizotinib, Brigatinib, or Alectinib). Ten patients received combination therapies, including EGFR-TKIs plus chemotherapy (n = 7) or EGFR-TKIs plus anti-angiogenic therapy (n = 3). Five patients underwent chemotherapy combined with immunotherapy (n = 4) or anti-angiogenic therapy (n = 1). Additionally, some patients received palliative care, including local radiotherapy. The overall response rate (ORR) and disease control rate (DCR) for first-line treatment were 63.10% and 97.62%, respectively. The median progression-free survival (PFSa) after first-line treatment was 16.70 months (95% CI: 13.04–20.37 months). No significant difference in PFSa was observed between patients transformed to SCLC and those transformed to other NSCLC subtypes (11.97 months vs. 9.80 months; P = 0.955; Supplementary Fig. 1). The median overall survival (OSa) after first-line treatment was 34.55 months (95% CI: 30.01–39.09 months), with no significant difference between the two transformation groups (35.63 months vs. 37.37 months; P = 0.441; Supplementary Fig. 1). First-line treatment and its efficacy in patients after HT The treatment details after HT are summarized in Supplementary Table 3. During follow-up, 27 patients survived, while 51 died (60.71%). Among the treated patients, 30 received chemotherapy alone, with etoposide–platinum being the most common regimen (n = 23). Other regimens included pemetrexed plus platinum (n = 4), nab-paclitaxel plus platinum (n = 2), and antibody–drug conjugates (n = 1). Additionally, 20 patients received a combination of TKIs and chemotherapy, and 25 received chemotherapy combined with immunotherapy or anti-angiogenic therapy. Nine patients underwent palliative care, including local radiotherapy. The ORR and DCR for the entire cohort were 26.19% (22/76) and 63.10% (53/84), respectively. The median PFS1 after first-line treatment post-HT was 9.7 months (95% CI: 8.4–10.9 months), and the median OS was 34.5 months (95% CI: 23.5–45.6 months). In our analysis of three treatment groups—CTK, CAI, and PC—we observed the following outcomes. For PFS1, the median values were 4.33 months in the CTK group, 4.27 months in the CAI group, and 2.03 months in the PC group. Despite a numerical trend favoring the CTK group, no statistically significant differences were observed (P = 0.259, Fig. 1A). For OS1, the medians were 15.40 months (CTK), 11.23 months (CAI), and 4.37 months (PC). Here, the differences were statistically significant (P = 0.003, Fig. 1B), with the CTK group demonstrating superior OS1 compared to the other two groups. However, for OS, the medians were 43.27 months (CTK), 40.47 months (CAI), and 39.73 months (PC), with no statistically significant differences (P = 0.139, Fig. 1C). These findings suggest that while the CTK regimen may extend OS1, it does not significantly improve PFS1 or OS when compared to the CAI and PC regimens. Progression-free survival (PFS1) and overall survival (OS1/OS) outcomes by treatment regimens and chemotherapy stratification. (A) PFS1 in patients receiving CTK (chemotherapy + TKIs), CAI (chemotherapy + immunotherapy/anti-angiogenics), or PC (palliative care); (B) OS1 in patients receiving CTK (chemotherapy + TKIs), CAI (chemotherapy + immunotherapy/anti-angiogenics), or PC (palliative care); (C) OS in patients receiving CTK (chemotherapy + TKIs), CAI (chemotherapy + immunotherapy/anti-angiogenics), or PC (palliative care); (D) PFS1 stratified by chemotherapy regimens: etoposide–carboplatin (EC), etoposide–platinum (EP), or other regimens; (E) OS1 stratified by chemotherapy regimens: etoposide–carboplatin (EC), etoposide–platinum (EP), or other regimens; (F) OS stratified by chemotherapy regimens: etoposide–carboplatin (EC), etoposide–platinum (EP), or other regimens Further stratification of chemotherapy regimens into three groups—etoposide-carboplatin (EC), etoposide–platinum (EP), and other regimens—revealed similar trends. For PFS1, the EP group showed numerically longer survival than the EC and other groups, but the differences were not statistically significant (P = 0.452, Fig. 1D). Similarly, for OS1 and OS, the EC group exhibited numerically higher median survival times compared to the EP and other groups, yet these differences also lacked statistical significance (P = 0.904 and P = 0.146, respectively, Fig. 1E, F). Collectively, these results indicate that, while certain chemotherapy regimens show numerical advantages, none achieved statistical significance, underscoring the need for further investigation to clarify their impact on patient outcomes. In summary, the CTK regimen demonstrated potential benefits in extending OS1 but did not significantly improve PFS1 or OS compared to the CAI and PC regimens. Larger studies or investigations in different patient populations are warranted to validate these findings and elucidate the underlying mechanisms driving these outcomes. Treatment and its efficacy in patients after HT-SCLC In our analysis of a cohort of SCLC patients, we evaluated the outcomes across three treatment groups: CAI, CTK, and PC. For PFS1, the median PFS1 was 5.17 months in the CTK Group, 4.20 months in the CAI Group, and 2.03 months in the PC Group. The differences in PFS1 among the three groups were statistically significant (P = 0.014, Fig. 2A), indicating that the CTK regimen provided superior progression-free survival compared to both the CAI and PC regimens. Regarding OS1, the median survival times were 15.40 months in the CTK Group, 11.07 months in the CAI Group, and 4.37 months in the PC Group. Notably, the differences in OS1 among the three groups reached statistical significance (P < 0.05, Fig. 2B), with the CTK group demonstrating significantly longer OS1 than bo; th the CAI and PC groups. This finding highlights the potential of the CTK regimen to offer substantial survival benefits in this patient population. For OS, the median survival times were 43.57 months in the CTK Group, 42.80 months in the CAI Group, and 23.83 months in the PC Group. Although there was a trend favoring the CTK and CAI groups over the PC group, the differences in OS among the three groups did not achieve statistical significance (P = 0.094, Fig. 2C). This suggests that while both the CTK and CAI regimens may provide some benefit in terms of overall survival, these differences are not robust enough to be distinguished from each other within the limitations of this study. Progression-free survival (PFS1) and overall survival (OS1/OS) outcomes by treatment regimens in small cell lung cancer (SCLC) patients. (A) PFS1 in patients receiving CTK (chemotherapy + TKIs), CAI (chemotherapy + immunotherapy/anti-angiogenics), or PC (palliative care); (B) OS1 in patients receiving CTK (chemotherapy + TKIs), CAI (chemotherapy + immunotherapy/anti-angiogenics), or PC (palliative care); (C) OS in patients receiving CTK (chemotherapy + TKIs), CAI (chemotherapy + immunotherapy/anti-angiogenics), or PC (palliative care) Second-line treatment and its efficacy in patients after HT Among 46 patients who received second-line treatment after HT, 12 underwent chemotherapy alone, 9 received anti-angiogenic therapy alone, and 8 received chemotherapy combined with immunotherapy or anti-angiogenic therapy. Twelve patients received TKI alone or in combination, and 2 received palliative care, including local radiotherapy. One patient achieved a partial response (PR) with a four-drug combination (ABCP: albumin-bound paclitaxel, bevacizumab, carboplatin, and paclitaxel), demonstrating a PFS of 7 months. In our analysis of second-line treatment outcomes, we compared the TKI-Included (TI) Group with the TKI-Excluded (TE) Group. For progression-free survival in the second line (PFS2), the median PFS2 was significantly longer in the TI Group (6.70 months) compared to the TE Group (2.87 months), with a statistically significant difference (P = 0.016, Supplementary Fig. 2A). However, for overall survival (OS), although the TI Group showed a numerically longer median OS (46.93 months) compared to the TE Group (42.80 months), this difference did not reach statistical significance (P = 0.163, Supplementary Fig. 2B). These findings indicate that, while including TKI in second-line treatment significantly improves PFS2, it does not significantly impact overall survival within the studied population. In our analysis of second-line treatment outcomes, we compared the Anti-Angio-Included (AI) Group with the Anti-Angio-Excluded (AE) Group. For progression-free survival in the second line (PFS2), the median PFS2 was numerically longer in the AE Group (3.97 months) compared to the AI Group (2.77 months), but this difference did not reach statistical significance (P = 0.322, Supplementary Fig. 2C). Similarly, for overall survival (OS), although the AI Group showed a numerically longer median OS (43.27 months) compared to the AE Group (39.60 months), this difference also did not achieve statistical significance (P = 0.322, Supplementary Fig. 2D). These findings suggest that the inclusion of anti-angiogenic therapy in second-line treatment does not significantly impact either PFS2 or OS within the studied population. Treatment and its efficacy in EGFR uncommon mutation patients after HT A total of 11 patients with uncommon EGFR mutations were included in this study (Table 2), comprising three patients with EGFR exon 19 p.L747_T751del, three patients with EGFR exon 21 L861Q mutation, one patient each with EGFR exon 19 p.L747_A750del, EGFR exon 19 p.E745_A750del, and EGFR exon 19 p.T751_I759del, as well as two patients with compound mutations: one with EGFR exon 18 [G719S + E709K] and another with EGFR exon 18 [G719S + S768I]. All patients received EGFR-TKI treatment prior to transformation. Among them, eight patients experienced small cell transformation, two patients experienced squamous cell transformation, and one patient developed neuroendocrine carcinoma, with a median time to transformation of 9.37 months. Regarding subsequent treatments post-transformation, three patients received immunotherapy combined with other treatments (one with immunotherapy plus anti-angiogenic therapy and two with immunotherapy plus chemotherapy), four patients underwent chemotherapy alone, one patient received chemotherapy combined with anti-angiogenic therapy, and two patients restarted TKI treatment. The ORR was 9.1%, and the DCR was 45.5%. The median PFS1 was 2.667 months, and the median OS1 was 10.967 months. When comparing outcomes between patients who did not receive immunotherapy versus those who received immunotherapy-containing regimens, we observed PFS1 of 1.83 months vs. 3.10 months (P = 0.007, Fig. 3A), OS1 of 5.93 months vs. 10.97 months (P = 0.500, Fig. 3B), OS of 15.30 months vs. 18.93 months (P = 0.390, Fig. 3B), ORR of 0% vs. 33.3% (P = 0.273), and DCR of 25% vs. 100% (P = 0.061). Table 2Characterization of histologically transformed advanced-NSCLC cases with uncommon EGFR mutationsMutation statusAge/genderSmoking historyBMTime to HT, mHistological typeFirst-line treatment post-HTLines of treatment post-HTORPFS, mOS,mPre-HTPost-HTEGFR 19 p.L747_T751del (Patient 1)57/MNoNo13.0LUADSCLCEC1SD1.82.4EGFR 19 p.L747_T751del (Patient 2)60/FNoNo17.5LUADSCLCTKI + EC2PD2.721.8EGFR 19 p.L747_T751del(Patient 3)83/MNoNo4.4LUADLSCCTKI1PD0.93.6EGFR 19 p.L747_A750del73/FNoNo5.2LUADNEC-NOSIAT1PD0.84.0EGFR 19 p.E745_A750del35/MNoNo40.2LUADLSCCIAT1SD42.743.2EGFR 19 p.T751_I759del69/FNoNo1.7LUADSCLCEC2SD2.812.2EGFR 18[G719S + E709K]52/MNoNo17.3LUADSCLCEC2PD0.514.7EGFR 18[ G719S + S768I]66/FNoNo14.6LUADSCLCNab-paclitaxel1PD0.61.3EGFR 21 L861Q(Patient 1)61/MYesNo7.1LUADSCLCCIT1PR3.08.3EGFR 21 L861Q(Patient 2)70/MYesNo9.4LUADSCLCPemetrexed2PD1.85.9EGFR 21 L861Q(Patient 3)77/MYesNo8.0LUADSCLCCIT1SD3.111.0Abbreviations: BM, brain metastasis; HT, histological transformation; OR, objective response; m, month; PFS, progression-free survival; OS, overall survival; LUAD, lung adenocarcinoma; LSCC: lung squamous cell carcinoma; EC, etoposide plus cisplatin; TKI, tyrosine kinase inhibitor; NEC-NOS, neuroendocrine carcinoma not otherwise specified; SD, stable disease; PD, disease progression; PR, partial relief; IAT: immunotherapy and anti-angiogenic therapy; CIT: chemotherapy and immunotherapy Characterization of histologically transformed advanced-NSCLC cases with uncommon EGFR mutations EGFR 19 p.L747_T751del (Patient 3) EGFR 18 [G719S + E709K] EGFR 18 [ G719S + S768I] EGFR 21 L861Q (Patient 1) EGFR 21 L861Q (Patient 2) EGFR 21 L861Q (Patient 3) Abbreviations: BM, brain metastasis; HT, histological transformation; OR, objective response; m, month; PFS, progression-free survival; OS, overall survival; LUAD, lung adenocarcinoma; LSCC: lung squamous cell carcinoma; EC, etoposide plus cisplatin; TKI, tyrosine kinase inhibitor; NEC-NOS, neuroendocrine carcinoma not otherwise specified; SD, stable disease; PD, disease progression; PR, partial relief; IAT: immunotherapy and anti-angiogenic therapy; CIT: chemotherapy and immunotherapy Progression-free survival (PFS1) and overall survival (OS1/OS) outcomes by immunotherapy use in patients with uncommon EGFR mutations. (A) PFS1 in patients receiving non-immunotherapy regimens vs. immunotherapy-containing regimens; (B) OS1 in patients receiving non-immunotherapy regimens vs. immunotherapy-containing regimens; (C) OS in patients receiving non-immunotherapy regimens vs. immunotherapy-containing regimens Tumor microenvironment profiling Multiplex immunofluorescence analysis of post-transformation specimens revealed distinct immune microenvironment profiles between classical EGFR-sensitive mutants (ex19del/L858R, n = 6) and uncommon EGFR mutants (n = 7). Patients with uncommon EGFR mutations exhibited a paradoxical immune landscape characterized by both heightened cytotoxic potential and enhanced immunosuppressive regulation. While CD8 + T cell infiltration was significantly elevated in uncommon mutants (median density: 128.5 vs. 64.2 cells/mm² in classical mutants, p = 0.028), this effector response was counterbalanced by a markedly increased Foxp3 + CD25+/CD8 + regulatory T cell ratio (0.41 vs. 0.19, p = 0.039), indicative of a Treg-dominated immunosuppressive milieu. Notably, PD-L1 expression was broadly low across both cohorts (median: <0.1 in 85.7% of uncommon mutants vs. 83.3% in classical mutants), with adaptive immune resistance (CD8+/PD-L1 + co-expression) observed in only 1/7 uncommon mutants. CD68 + macrophage infiltration showed no intergroup difference (median: 92.4 vs. 85.7 cells/mm², p = 0.612), though uncommon mutants displayed higher CD68+/PD-L1 + co-localization (median: 0.0021 vs. 0.0019, p = 0.086), suggesting macrophage-mediated PD-L1 suppression. These immunophenotypic distinctions aligned with clinical outcomes: uncommon mutants receiving immunotherapy-containing regimens achieved superior progression-free survival versus non-immunotherapy approaches (3.10 vs. 1.83 months, p = 0.007), despite comparable overall survival (10.97 vs. 5.93 months, p = 0.500). The discordance between enhanced CD8 + infiltration and limited clinical benefit (ORR: 9.1%, DCR: 45.5%) underscores the dominance of immunosuppressive mechanisms (Tregs/macrophages) in shaping therapeutic resistance, even within immunogenically active uncommon EGFR-mutant tumors (Fig. 4). Tumor immune microenvironment profiling in EGFR-mutant tumors via multiplex immunofluorescence. (A) Representative multiplex immunofluorescence images of an EGFR exon 19 deletion case; (B) representative multiplex immunofluorescence images of an EGFR exon 21 L861Q mutation case Discussion The phenomenon of HT in NSCLC, particularly following treatment with EGFR-TKIs, represents a significant clinical challenge. Our study aims to provide novel insights into the incidence, clinical characteristics, and therapeutic outcomes associated with HT, emphasizing its complexity and heterogeneity. By analyzing one of the largest cohorts to date, we identified several key findings that align with and extend prior research while addressing important gaps in knowledge. Previous studies have reported an HT incidence ranging from 3% to 15% in EGFR-mutant NSCLC patients treated with TKIs [6–8]. In our cohort, the HT incidence was 1.35%, which is lower than previously reported figures. This discrepancy may be attributed to differences in patient selection criteria, biopsy availability, or molecular profiling methods. For instance, earlier studies often relied on clinical suspicion rather than confirmatory biopsies for HT diagnosis [7], potentially overestimating its occurrence. Furthermore, our analysis included a broader range of molecular alterations beyond common EGFR mutations, capturing rare variants that might exhibit distinct transformation patterns. The predominance of SCLC transformation (73.81%) in our study corroborates findings from other investigations [16, 17, 24]. The enrichment of TP53 and RB1 co-mutations in SCLC-transformed cases is consistent with their established role in neuroendocrine differentiation and therapeutic resistance [12, 25]. However, unlike previous reports focusing predominantly on classical EGFR mutations (exon 19 deletions and L858R), our findings revealed unique immune microenvironment profiles in uncommon EGFR mutants, characterized by heightened cytotoxic potential counterbalanced by immunosuppressive regulation. These observations could indicate that rare mutations may warrant tailored therapeutic approaches beyond standard TKI-based regimens. Our results suggest a potential survival advantage of CTK in our cohort after HT compared to chemotherapy combined with CAI or PC. The median OS1 for the CTK group was observed to be longer (15.40 months vs. 11.23 months for CAI and 4.37 months for PC; P = 0.003). While this finding aligns with emerging evidence supporting TKI continuation post-HT [26, 27], the lack of significant differences in PFS1 underscores the need for optimized combination strategies. Recent trials exploring next-generation TKIs in combination with platinum-based chemotherapy have shown promise but remain limited by small sample sizes and heterogeneous populations [28, 29]. In second-line treatment, incorporating TKIs resulted in a statistically significant prolongation of PFS2 (6.70 vs. 2.87 months; P = 0.016). However, no significant improvement in overall survival was observed, implying that additional mechanisms of resistance may emerge during later disease stages. This observation echoes findings from preclinical models demonstrating adaptive resistance pathways activated upon prolonged TKI exposure [30]. Consistent with Liu et al. [31], who identified TAP1 and IRF4 as predictive biomarkers for immunotherapy response in triple-negative breast cancer and highlighted the role of hypoxia and glycolysis in treatment resistance, our findings underscore the criticality of biomarker-driven patient stratification. In our cohort, the paradoxical immune profile of uncommon EGFR mutants – elevated cytotoxic T cells alongside heightened immunosuppression – mirrors the TME complexity reported by Liu et al., where activated immune pathways coexisted with resistance-related metabolic alterations. Furthermore, the limited OS benefit of immunotherapy in our study aligns with Liu et al.’s observation that aggravated hypoxia and glycolysis can undermine immunotherapeutic efficacy, suggesting that targeting these metabolic pathways alongside immune checkpoints may improve outcomes in HT patients with uncommon EGFR mutations. Multiplex immunofluorescence analysis revealed distinct immune microenvironment profiles between classical EGFR-sensitive mutants and uncommon EGFR mutants. Elevated CD8 + T-cell infiltration in uncommon mutants was accompanied by increased regulatory T-cell activity, creating a paradoxical landscape of enhanced cytotoxic potential yet robust immunosuppression. Notably, patients receiving immunotherapy-containing regimens achieved superior PFS (3.10 vs. 1.83 months; P = 0.007), despite comparable OS. This suggests that while immunotherapy can delay disease progression, its long-term efficacy may be limited by persistent immunosuppressive mechanisms, such as Tregs and macrophage-mediated PD-L1 suppression. Consistent with Liu et al., who identified TAP1 and IRF4 as predictive biomarkers for immunotherapy response in triple-negative breast cancer and highlighted the role of hypoxia and glycolysis in treatment resistance, our findings underscore the criticality of biomarker-driven patient stratification. In our cohort, the paradoxical immune profile of uncommon EGFR mutants – elevated cytotoxic T cells alongside heightened immunosuppression – mirrors the TME complexity reported by Liu et al., where activated immune pathways coexisted with resistance-related metabolic alterations. Furthermore, the limited OS benefit of immunotherapy in our study aligns with Liu et al.’s observation that aggravated hypoxia and glycolysis can undermine immunotherapeutic efficacy, suggesting that targeting these metabolic pathways alongside immune checkpoints may improve outcomes in HT patients with uncommon EGFR mutations. The discordance highlights the importance of biomarker-driven stratification to identify subsets most likely to benefit from immune checkpoint inhibition. Future studies should explore combinatorial approaches targeting both effector and suppressive arms of the immune system in rare EGFR-mutant tumors. Additionally, Oyoshi et al. [32]. (2023) demonstrated that radiotherapy induces infiltration of PD-L1-positive myeloid cells expressing multiple immunosuppressive (SIRPA, IDO1) and protumor (CXCL8, IL-6) genes, indicating that dual targeting of PD-L1 with these molecules could enhance therapeutic responses. While our study focused on TKI-induced HT rather than radiotherapy, the identification of similar immunosuppressive pathways in the TME of uncommon EGFR mutants suggests that combinatorial immunotherapies targeting PD-L1 alongside SIRPA or IDO1 may overcome resistance in this subset. This is further supported by Liu et al.’s (2024) emphasis on multiomics biomarker analysis to guide combination strategies, reinforcing that a single-agent approach is insufficient to address the complex TME dynamics in treatment-refractory cancers. Despite its strengths, our study has several limitations that warrant acknowledgment. First, the retrospective design introduces inherent biases, including potential selection bias due to differential access to re-biopsy confirmation. Second, the sample sizes for key subgroup analyses remain limited. Due to the small sample size of subgroups (e.g., ALK-positive cases, n = 6; uncommon EGFR mutations, n = 11), the findings should be interpreted as exploratory and hypothesis-generating rather than definitive. Large-scale prospective studies are needed to validate the clinical implications of these subgroups. Furthermore, the CAI group is heterogeneous in its therapeutic mechanisms, and larger future studies are needed to evaluate the efficacy of these strategies separately. Consequently, the statistical comparisons and efficacy outcomes presented for these subgroups should be viewed as exploratory and hypothesis-generating. Third, longitudinal assessment of tumor evolution and immune dynamics was constrained by available tissue samples, precluding detailed characterization of temporal changes post-HT. Lastly, the lack of standardized definitions for HT across studies hampers direct comparisons, underscoring the need for consensus guidelines in future research. Due to the small sample size of subgroups (e.g., ALK-positive cases, n = 6; uncommon EGFR mutations, n = 11), the findings should be interpreted as exploratory and hypothesis-generating rather than definitive. Large-scale prospective studies are needed to validate the clinical implications of these subgroups. Conclusion HT in advanced NSCLC poses a significant clinical challenge due to its association with acquired resistance and altered treatment responses. Our study demonstrates the feasibility of personalized treatment strategies based on histologic subtype, molecular alterations, and immune contexture. Chemotherapy combined with TKIs emerges as the most effective regimen for improving overall survival after HT, while second-line inclusion of TKIs prolongs progression-free survival. The distinct, paradoxical immune microenvironment profiles in transformed tumors, particularly those with uncommon EGFR mutations—characterized by a “hot but suppressed” state—highlight the necessity for tailored, combinatorial approaches. Distinct immune microenvironment profiles in rare EGFR mutants underscore the necessity for tailored approaches addressing both cytotoxic and suppressive pathways. While these findings provide valuable insights, further prospective studies are required to validate conclusions and refine therapeutic paradigms. Addressing the limitations of current knowledge through large-scale multicenter collaborations and innovative trial designs will be essential for advancing precision medicine in this complex population. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Abbreviations Epidermal growth factor receptor Anaplastic lymphoma kinase Non-small cell lung cancer Histologic transformation Tumor microenvironment Chemotherapy combined with tyrosine kinase inhibitors Chemotherapy combined with anti-angiogenic/immunotherapy Palliative care Overall survival Progression-free survival Small cell lung cancer Tyrosine kinase inhibitor CD8-positive T cells Regulatory T cell Forkhead box P3-positive American Joint Committee on Cancer Tumor, node, metastasis World Health Organization Formalin-fixed paraffin-embedded Next-generation sequencing Polymerase chain reaction Response Evaluation Criteria in Solid Tumors 4’,6-diamidino-2-phenylindole Ethylenediaminetetraacetic acid Hazard ratio Confidence interval Retinoblastoma 1 Tumor Protein 53 Programmed death-ligand 1 Etoposide–carboplatin Etoposide–platinum Overall response rate Disease control rate Partial response Complete response Stable disease Progressive disease Neuroendocrine carcinoma not otherwise specified Performance status Immune checkpoint blockade Monoclonal antibody Author contributions Kaibo Ding: Conceptualization, data curation, software, methodology, writing—original draft. Xinyue Li: Data curation, methodology. Yanjun Xu: Visualization, supervision, writing—review and editing. Funding This study was supported by the Natural Scientific Foundation of Zhejiang Province, China (Grant No. LTGY23H160007). Data availability The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate Written informed consent for the use of patient specimens was obtained from each participant. The study protocols were approved by the Ethics Committee of Zhejiang Cancer Hospital (Hangzhou, China). Consent for publication Not applicable. Competing interests The authors declared no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Kaibo Ding and Xinyue Li contributed equally as co-authors of this article. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Data Availability Statement The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.