Metabolic tumor volume on ¹⁸F-fluorodeoxyglucose uptake as prognostic marker for osimertinib in patients with non-small cell lung cancer harboring sensitive EGFR mutation

Introduction

Epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) are standard treatments for patients with advanced non-small cell lung cancer (NSCLC) harboring EGFR mutations. Among the EGFR-TKIs, osimertinib, a 3^rd generation TKI, is a potent drug that improves survival compared with 1^st or 2^nd generation EGFR-TKIs (1).

Approximately 70% of patients with sensitizing EGFR mutations experience marked tumor shrinkage; however, established markers for predicting therapeutic response and outcome after EGFR-TKI treatment remain unclear.

Fluorine-18 fluorodeoxyglucose (¹⁸F-FDG) positron emission tomography (PET) is a radiological modality used to differentiate malignant from benign tumors by measuring their metabolic activity. Many studies have reported that ¹⁸F-FDG uptake in tumor lesions can serve as a prognostic marker in patients with NSCLC (2). Several reports have described that a low primary maximum standardized uptake value (SUV_max) is associated with EGFR mutations and may predict therapeutic response and prognosis after EGFR-TKIs treatment in EGFR-mutated lung adenocarcinoma (AC) (3-6). However, previous studies have not included osimertinib as an EGFR-TKI when assessing ¹⁸F-FDG accumulation. Anai et al. reported that high primary lesion SUV was significantly correlated with short progression-free survival (PFS) after first-line osimertinib treatment in 74 patients with EGFR mutation-positive NSCLC (7). Moreover, Leonetti et al. reported that an early metabolic response led to improved PFS after first-line (n=63) or second-line (n=9) osimertinib treatment in patients with advanced EGFR-mutated NSCLC (8). Based on previous PET studies, metabolic tumor volume (MTV) and total lesion glycolysis (TLG) have been described as significant predictors of outcomes after systemic treatment in patients with advanced NSCLC (9,10). MTV reflects the total tumor volume with activating cancer cells and is thus associated with therapeutic resistance, progression of disease, and performance status (PS). Tumor tissues with EGFR mutations reflect heterogeneity, including sensitizing, uncommon, and compound mutations. Measuring this heterogeneity using limited parameters such as SUV_max may be challenging. However, the relationship between EGFR-TKI treatment, MTV, TLG, and ¹⁸F-FDG uptake in patients with EGFR mutation-positive NSCLC remains poorly understood. Based on this background, we retrospectively conducted a study to evaluate the therapeutic efficacy and outcomes of osimertinib in patients with advanced NSCLC harboring sensitive EGFR mutations. We present this article in accordance with the STROBE reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-787/rc).

Methods

Patients

This retrospective study included patients with advanced NSCLC harboring sensitizing EGFR mutations who received first-line osimertinib treatment at Saitama Medical University, International Medical Center between June 2018 and August 2021. A total of 103 patients were initially identified; of these, 68 patients who underwent ¹⁸F-FDG PET before initiation of osimertinib were included in the final analysis.

Thirty-five patients were excluded for the following reasons: 17 underwent PET more than 8 weeks before osimertinib initiation; 17 underwent PET at outside institutions or lacked PET imaging data; and one patient had incomplete clinical data. Relevant clinical information was retrospectively collected from the electronic medical records at Saitama Medical University International Medical Center.

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Ethics Committee of International Medical Center, Saitama Medical University (approval No. koku 2025-022). A waiver of informed consent was granted due to the retrospective study design (11).

Treatment and evaluation

All patients received oral osimertinib (80 mg/day) daily, as previously described (1). Each chief physician determined the timing and necessity of physical examinations, complete blood counts, biochemical tests, and adverse event evaluations. Toxicities were assessed and graded according to Common Terminology Criteria for Adverse Events (CTCAE), version 4.0 (12). Tumor response was evaluated in accordance with the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 (13).

PET imaging and data analysis

Patients were instructed to fast for at least 6 hours prior to the examination. At the beginning of the examination, ¹⁸F-FDG was administered intravenously. Three-dimensional (3D) PET/CT data acquisition was initiated at 60 minutes after FDG injection using either a Biograph 16 or Biograph mCT 600 scanner (Siemens Healthineers K.K., Tokyo, Japan). The number of bed positions, which was typically eight, was determined based on the imaging range. Attenuation-corrected transverse PET images were reconstructed into 168×168 matrices with a slice thickness of 2.00 mm using an ordered subset expectation maximization (OSEM) algorithm incorporating a point-spread function.

We used Syngo. via (Siemens Healthineers Co. Ltd., Tokyo, Japan) on a Windows workstation to semi-automatically calculate the SUV_max and peak SUV (SUV_peak). For ¹⁸F-FDG PET, the MTV and TLG, defined as the MTV multiplied by the mean SUV (SUV_mean) for each lesion, were calculated using SUV_threshold derived from the SUV of the liver volume of interest (VOI). Each threshold was defined as the average of 1.5 × SUV_mean plus 2 × the standard deviation of the liver SUV. These SUVthresholds were determined to be optimal for generating 3D VOIs that completely enclosed the entire tumor mass, with computed tomography (CT) images used as a reference. Regions of activity unrelated to tumors, including the kidneys, urinary tract, myocardium, and gastrointestinal tract, were manually excluded by a board-certified nuclear medicine physician based on the orientation provided. The liver VOI was automatically set using artificial intelligence (AI) analysis software.

Statistical analysis

Fisher’s exact test was used to examine the association between two categorical variables. Statistical significance was set at P<0.05. According to RECIST, patients were classified into four categories: complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD). The optimal cutoff values for ¹⁸F-FDG uptake parameters, SUV_max, SUV_peak, MTV, and TLG, were determined based on their median values: SUV_max =7.9, SUV_peak =6.7, MTV =85.5, and TLG =304.7. These values were used to differentiate patients between groups. PFS was defined as the time from initial treatment to disease progression or death, while OS was defined as the time from initial treatment to death from any cause. Survival curves were estimated using the Kaplan-Meier method, with median survival time (MST) reported for each group. Positive programmed death ligand-1 (PD-L1) expression was defined as >1% (PD-L1 >1%). Differences in survival distributions were assessed using log-rank tests. Univariate and multivariate analyses were performed using Cox proportional hazards models to calculate hazard ratios (HRs) with 95% confidence intervals (CIs). Variables included in the multivariate analysis were selected based on their clinical relevance, such as patient background characteristics and previous studies identifying the MTV as a significant prognostic factor (9,10). All statistical analyses were performed using GraphPad Prism software (v.10.0; GraphPad Software, San Diego, CA, USA) and JMP Pro 16.0 (SAS Institute Inc., Cary, NC, USA).

Results

Patient characteristics

Patient characteristics based on ¹⁸F-FDG accumulation are summarized in Table 1. This study included 68 patients (30 males and 38 females; median age, 72 years; age range, 45–88 years). Patients aged >70 years had significantly higher SUV_max. Thirty-two patients (47.1%) had a history of smoking. The PS was 0–1 in 55 patients (80.9%), two in six patients (8.8%), and 3–4 in seven patients (10.3%). Histology included AC in 66 patients (97.1%) and non-adenocarcinoma (non-AC) in two patients (2.9%). Sixty-five patients were at clinical stage of IV, and three patients were at stage of III. In addition to 28 patients with exon 19 deletion and 37 with L858R, there was one patient each with an exon 20 insertion, G719X, and a triple mutation including L858R, S768I, and T790M.

Metastases to the brain, bones, and liver were observed in 24 (35.3%), 34 (50.0%), and 59 (86.8%) patients, respectively. Bone metastasis was significantly associated with high MTV and TLG values, while liver metastasis was associated with elevated values across all four parameters (SUV_max, SUV_peak, MTV, and TLG). Regarding PD-L1 expression, 33 patients (48.5%) were positive, 31 patients (45.6%) were negative, and four patients (5.9%) had unknown status. Positive PD-L1 expression was significantly associated with higher SUV_max and SUV_peak values than those with negative PD-L1 expression. A PR or CR was observed in 42 patients (61.8%), SD or PD in 21 patients (30.9%), and response status was unknown in four patients (5.9%). Grade 3 or higher adverse events occurred in 15 patients (22.1%).

Therapeutic efficacy and adverse events

Table 2 presents a comparison of tumor response (CR or PR versus SD or PD) and disease control (non-PD versus PD) according to continuous variables derived from ¹⁸F-FDG accumulation. A significant difference in tumor response was observed in relation to MTV and TLG values from ¹⁸F-FDG uptake; however, not such difference was found in the disease control group (Table 2). Figure S1 displays waterfall plots illustrating tumor shrinkage based on the MTV derived from ¹⁸F-FDG uptake.

Table 3 summarizes the comparison of adverse events according to continuous variables related to ¹⁸F-FDG uptake. No significant differences were observed in the incidence of grade 3/4 adverse events or skin toxicity. However, high mean values of SUV_max and SUV_peak from ¹⁸F-FDG uptake were significantly associated with the occurrence of lung toxicity induced by osimertinib (Table 3).

Survival analysis based on ¹⁸F-FDG PET accumulation

The follow-up period was 757 days (range, 57–2,275 days). The median PFS and OS after the initial treatment were 382 and 757 days, respectively. In the PFS analysis, disease progression was observed in 50 patients. Four patients were transitioned to the best supportive care because of adverse events. Drug changes occurred in four patients: three due to adverse events and one due to the use of alternative medicine. Two patients died as a result of adverse events, and one patient died from an unrelated disease. Osimertinib treatment was ongoing in the remaining seven patients. Regarding OS, a total of 55 patients died during the follow-up period, including those who subsequently received further treatments. The Kaplan-Meier curves for PFS and OS according to ¹⁸F-FDG uptake are presented in Figures 1,2, respectively. PFS and OS differed significantly between the two groups based on MTV (PFS: HR =1.91, 95% CI: 1.14–3.22, P<0.01; OS: HR =2.00, 95% CI: 1.16–3.43, P<0.01) and TLG (PFS: HR =1.70, 95% CI: 1.01–2.84, P=0.03; OS: HR =1.72, 95% CI: 1.00–3.00, P=0.04). In contrast, no significant differences in PFS or OS were observed between the two groups based on SUV_max or SUV_peak.

Figure 1 Kaplan-Meier curves for PFS according to ¹⁸F-FDG uptake. Kaplan-Meier curves of PFS according to SUV_max (A), SUV_peak (B), MTV (C), and TLG (D) on ¹⁸F-FDG uptake. Although there was not significant difference in the PFS between high and low SUV_max (A), and between high and low SUV_peak (B), high MTV (C) and TLG (D) were significantly associated with shorter prognosis. CI, confidence interval; FDG, fluorodeoxyglucose; HR, hazard ratio; MTV, metabolic tumor volume; PFS, progression-free survival; SUV_max, maximum of standardized uptake value; SUV_peak, peak of standardized uptake value; TLG, total lesion glycolysis.

Figure 2 Kaplan-Meier curves for OS according to ¹⁸F-FDG uptake. Kaplan-Meier curves of OS according to SUV_max (A), SUV_peak (B), MTV (C), and TLG (D) on ¹⁸F-FDG uptake. There was not significant difference in the OS between high and low SUV_max (A), and between high and low SUV_peak (B), but, high MTV (C) and TLG (D) were significantly associated with worse prognosis. CI, confidence interval; FDG, fluorodeoxyglucose; HR, hazard ratio; MTV, metabolic tumor volume; OS, overall survival; SUV_max, maximum of standardized uptake value; SUV_peak, peak of standardized uptake value; TLG, total lesion glycolysis.

Univariate and multivariate analyses

Univariate and multivariate analyses were performed for all patients. Univariate analysis identified PS and MTV as significant predictors of PFS and OS, respectively (Table 4). Variables with P<0.05 in univariate log-rank tests were included in the multivariate analysis. Additionally, clinically relevant factors such as age and EGFR mutation status were incorporated, regardless of their univariate significance. The survival analysis according to the EGFR mutation status such as del 19 and L858R was listed in Figures S2,S3. No significantly significant difference in the PFS and OS was observed in the patients with del 19 EGFR mutation (Figure S1). In the patients with L858R, high MTV was identified as significant worse OS (Figure S1).

In the multivariate analysis for PFS, MTV was identified as a significant prognostic factor. Although PS was selected as a significant predictor in the univariate analysis, it was not significant in the multivariate analysis (Table 4). Multivariate analysis using TLG instead of MTV was listed in Table S1, and TLG was not selected as an independent factor predicting worse outcome.

Discussion

Our study successfully identified metabolic tumor activity as a potential prognostic marker after osimertinib treatment in patients with EGFR mutation-positive NSCLC. Consistent with our previous study (9,10), higher MTV, unlike SUV_max, was significantly associated with shorter PFS and OS, suggesting that the tumor volume consisting of malignant cells harboring activating mutations is involved in the mechanism of resistance to EGFR-TKIs. Currently, measuring heterogeneity related to varying mutation burdens using visual modalities remains impossible. Among the different radiological modalities, PET is more suitable than morphological devices, such as CT, for delineating the extent of activation of malignant cells. Moreover, we found that measurement of the local area using SUV_max was markedly limited in predicting outcomes after osimertinib treatment. Our study suggests that MTV based on ¹⁸F-FDG uptake is limited in predicting disease control and adverse events caused by osimertinib monotherapy, whereas low SUV_max and SUV_peak are closely associated with lung toxicity.

Recently, combined molecular targeted therapies, such as the FLAURA2 and MARIPOSA regimens, have demonstrated a significant survival benefit compared to osimertinib monotherapy (14,15). The FLAURA2 study reported that PFS was significantly longer in patients treated with osimertinib combined with chemotherapy than in those treated with osimertinib monotherapy (24.0 vs. 15.3 months, P<0.001) (14). MARIPOSA study described that the median PFS was significantly longer in patients receiving amivantamab combined with lazertinib than in those receiving osimertinib monotherapy (23.7 vs. 16.6 months, P<0.001) (15). Thus, additional treatment modalities alongside osimertinib are required to overcome resistance mechanisms in EGFR-mutated tumor cells. Recent data suggest that combination imaging and liquid biopsy biomarkers may guide treatment selections. For example, Leonetti et al. reported that PET parameters correlated with ctDNA variant allele frequency (8), and baseline MTV on ¹⁸F-FDG PET might therefore complement ctDNA as a surrogate biomarker for identifying patients who could benefit from novel combinations. Moreover, it remains unclear which treatment, chemotherapy or amivantamab, is better. Identifying an optimal biomarker to predict poor outcomes after osimertinib monotherapy, may facilitate the appropriate selection of combination therapies, such as those used in the FLUARA2 or MARIPOSA regimens. The association between PD-L1 expression and PET parameters is also of interest, as previous studies have reported conflicting results regarding the impact of high PD-L1 expression on outcomes with osimertinib (16-18). The novelty of our study lies in demonstrating that baseline MTV is a useful prognostic indicator for patients treated with osimertinib monotherapy and may serve as a promising prognostic factor for future evaluation in regimens such as FLAURA2 and MARIPOSA.

In the present study, the presence of liver metastases was closely associated with ¹⁸F-FDG accumulation as determined by SUV_max, MTV, and TLG, whereas the presence of bone metastases was significantly correlated with metabolic tumor activity as determined by MTV and TLG. Based on our findings and previous evidence (9,10), we believe that MTV or TLG more accurately reflect systemic disease expansion and tumor progression than SUV_max. In a previous study reporting osimertinib and ¹⁸F-FDG uptake, only SUV_max was assessed, without analysis of MTV or TLG (7). Although SUV_max was significantly associated with primary tumor size, the relationship between ¹⁸F-FDG uptake and systemic progression such as bone and liver metastases remains unclear (7). Moreover, higher SUV_max was closely related to shorter PFS, whereas no significant association was observed with OS (7). In contrast, our study revealed that the MTV was a significant predictor for short PFS and OS after osimertinib, while SUV_max was not. We hypothesized that the SUV reflects only partial tumor activity and not the entire tumor heterogeneity. Therefore, MTV may be more suitable for assessing tumors with heterogeneous resistance after osimertinib treatment than SUV_max. In our study, predicting tumor response and adverse events associated with osimertinib treatment based on ¹⁸F-FDG uptake levels proved difficult. We believe that the MTV, as measured by ⁸F-FDG uptake, plays a crucial role in prognostic prediction after osimertinib monotherapy.

This study has several limitations. First, the small sample size, which may have introduced bias into the results. In addition, the limited number of cases restricted the robustness of the multivariate analysis, and the findings should therefore be interpreted with caution. Moreover, because PET-CT could only be performed in patients who were sufficiently fit before treatment, our cohort may represent a relatively healthier subset of the population introducing a potential selection bias. Further studies with larger cohorts are needed to establish the prognostic significance of MTV as a predictive biomarker. Second, our study focused solely on the prognostic value of ¹⁸F-FDG uptake after osimertinib monotherapy. However, it may be essential to compare the prognostic utility of MTV across different treatment settings, such as first- or second-generation EGFR-TKIs and osimertinib combined with chemotherapy. Finally, we were unable to assess the specific types of tumor mutation burdens associated with MTV-based ¹⁸F-FDG uptake following osimertinib failure. We speculate that elevated MTV levels could be associated with a variety of tumor mutations. Thus, it is necessary to evaluate the relationship between MTV and genomic profiles after resistance to osimertinib. Further studies are warranted to evaluate the predictive role of ¹⁸F-FDG uptake based on comprehensive genomic profiles.

Conclusions

MTV is a significant marker for predicting outcomes after osimertinib monotherapy in patients with advanced NSCLC harboring EGFR mutations. Selecting first-line EGFR-TKIs according to the elevated MTV levels may help in avoiding resistance to osimertinib monotherapy.

Acknowledgments

The authors thank Ms. Kozue Watanabe for her assistance in preparing the manuscript and Dr. Ayako Shiono and Dr. Yu Miura for the collection of patients. The authors also thank Editage (www.editage.jp) for English language editing.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-787/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-787/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-787/prf

Funding: This work was supported by the JSPS Grant-in-Aid for Scientific Research C(No. 20K08118).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-787/coif). K.K. received the JSPS Grant-in-Aid for Scientific Research C (No. 20K08118), a speaker honorarium from Ono Pharmaceutical Company, Chugai Pharmaceutical, and AstraZeneca and research grants from AstraZeneca. A.M. received speaker honoraria from Chugai Pharmaceutical and AstraZeneca. O.Y. received speaker honoraria from Chugai Pharmaceutical and AstraZeneca. H.K. received research grants from Ono Pharmaceutical Company, Boehringer Ingelheim and Chugai Pharmaceutical; speaker honoraria from Ono Pharmaceutical, Chugai Pharmaceutical, AstraZeneca, Bristol-Myers, and MSD. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Ethics Committee of International Medical Center, Saitama Medical University (approval No. koku 2025-022). A waiver of informed consent was granted due to the retrospective study design.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Ramalingam SS, Vansteenkiste J, Planchard D, et al. Overall Survival with Osimertinib in Untreated, EGFR-Mutated Advanced NSCLC. N Engl J Med 2020;382:41-50. [Crossref] [PubMed]
2. Kaira K, Kuji I, Kagamu H. Value of (18)F-FDG-PET to predict PD-L1 expression and outcomes of PD-1 inhibition therapy in human cancers. Cancer Imaging 2021;21:11. [Crossref] [PubMed]
3. Ng KS, King Sun C, Boom Ting K, et al. Prognostic factors of EGFR-mutated metastatic adenocarcinoma of lung. Eur J Radiol 2020;123:108780. [Crossref] [PubMed]
4. Gao XC, Wei CH, Zhang RG, et al. (18)F-FDG PET/CT SUV(max) and serum CEA levels as predictors for EGFR mutation state in Chinese patients with non-small cell lung cancer. Oncol Lett 2020;20:61. [Crossref] [PubMed]
5. Shao D, Cheng Y, Yuan ZS, et al. Value of interim 18F-FDG PET/CT for predicting progression-free survival in stage IIIB/IV EGFR-mutant non-small-cell lung cancer patients with EGFR-TKI therapy. Lung Cancer 2020;149:137-43. [Crossref] [PubMed]
6. Agüloğlu N, Akyol M, Kömek H, et al. The Prognostic Value of (18)F-FDG PET/CT Metabolic Parameters in Predicting Treatment Response Before EGFR TKI Treatment in Patients with Advanced Lung Adenocarcinoma. Mol Imaging Radionucl Ther 2022;31:104-13. [Crossref] [PubMed]
7. Anai M, Inoue H, Saruwatari K, et al. Negative-predictive value of SUVmax for Ascertaining the efficacy of osimertinib in EGFR mutation-positive non-small cell lung cancer. Respir Investig 2024;62:1072-8. [Crossref] [PubMed]
8. Leonetti A, Cervati V, Minari R, et al. Liquid Biopsy and 18F-FDG PET/CT Derived Parameters as Predictive Factors of Osimertinib Treatment in Advanced EGFR-Mutated NSCLC. Clin Lung Cancer 2024;25:e436-e445.e9.
9. Hashimoto K, Kaira K, Imai H, et al. Prognostic Potential of Metabolic Activity on 18 F-FDG Accumulation in Advanced NSCLC Receiving Combining Chemotherapy Plus PD-1 Blockade. J Immunother 2022;45:349-57. [Crossref] [PubMed]
10. Hashimoto K, Kaira K, Yamaguchi O, et al. Visual Assessment of 18F-FDG Uptake on PET to Predict Survival Benefit to PD-1 Blockade in Non-Small Cell Lung Cancer. Clin Nucl Med 2022;47:108-16. [Crossref] [PubMed]
11. Eba J, Nakamura K. Overview of the ethical guidelines for medical and biological research involving human subjects in Japan. Jpn J Clin Oncol 2022;52:539-44. [Crossref] [PubMed]
12. National Cancer Institute. Common Terminology Criteria for Adverse Events (CTCAE) Version 4.0. National Institutes of Health; 2010. Available online: https://www.cancer.gov/research/resources/resource/210
13. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009;45:228-47. [Crossref] [PubMed]
14. Planchard D, Jänne PA, Cheng Y, et al. Osimertinib with or without Chemotherapy in EGFR-Mutated Advanced NSCLC. N Engl J Med 2023;389:1935-48. [Crossref] [PubMed]
15. Cho BC, Lu S, Felip E, et al. Amivantamab plus Lazertinib in Previously Untreated EGFR-Mutated Advanced NSCLC. N Engl J Med 2024;391:1486-98. [Crossref] [PubMed]
16. Brown H, Vansteenkiste J, Nakagawa K, et al. Programmed Cell Death Ligand 1 Expression in Untreated EGFR Mutated Advanced NSCLC and Response to Osimertinib Versus Comparator in FLAURA. J Thorac Oncol 2020;15:138-43. [Crossref] [PubMed]
17. Hamakawa Y, Agemi Y, Shiba A, et al. Association of PD-L1 tumor proportion score ≥20% with early resistance to osimertinib in patients with EGFR-mutated NSCLC. Cancer Med 2023;12:17788-97. [Crossref] [PubMed]
18. Yoshimura A, Yamada T, Okuma Y, et al. Impact of tumor programmed death ligand-1 expression on osimertinib efficacy in untreated EGFR-mutated advanced non-small cell lung cancer: a prospective observational study. Transl Lung Cancer Res 2021;10:3582-93. [Crossref] [PubMed]