Research ArticleCancer

A PET imaging approach for determining EGFR mutation status for improved lung cancer patient management

See allHide authors and affiliations

Science Translational Medicine  07 Mar 2018:
Vol. 10, Issue 431, eaan8840
DOI: 10.1126/scitranslmed.aan8840
  • Fig. 1 Molecular imaging of EGFR-activating mutation status for stratification of patients.

    (A) Schematic of epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) explaining the benefits to patients with non–small cell lung cancer (NSCLC) through a stratification strategy using 18F-MPG [N-(3-chloro-4-fluorophenyl)-7-(2(2-(2-(2-18F-fluoroethoxy) ethoxy) ethoxy) ethoxy)-6-methoxyquinazolin-4-amine] positron emission tomography (PET)/computed tomography (CT). 18F-MPG selectively binds to tumors expressing the EGFR-activating mutation, such that PET/CT can qualitatively and quantitatively reveal the EGFR mutation expression of tumors. The proposed concept can be eventually translated into clinical use. For patients with NSCLC, especially when the conventional EGFR mutation status test result is ambiguous and biopsies are not feasible or are inconclusive, 18F-MPG PET/CT can serve as a noninvasive, repeatable technique to comprehensively monitoring intra- and intertumor EGFR-activating mutation status in vivo. This strategy can be a valuable diagnostic tool to predict the EGFR-TKI sensitivity/resistance, survival, and guide precision EGFR-TKI treatment. (B) Synthesis of 18F-MPG; reagents and conditions: K18F/Kryptofix/dimethyl sulfoxide (DMSO), 120°C, 15 min, 23.79% radiochemical yield.

  • Fig. 2 In vitro and preclinical study with 18F-MPG.

    (A) Quantitative uptake of 18F-MPG in four NSCLC cell lines and pretreatment with gefitinib. Data are means ± SD of three independent experiments. **P < 0.01 versus line-matched 18F-MPG uptake in HCC827 cells, Student’s t test. (B) The autoradiography and the same Western blot membrane stained with E746-A750–specific antibody. A single 18F-labeled protein band corresponds to the predominant band of ~175 kDa. Graphs next to the blots show densitometric quantification of autoradiogram band normalized to background. ***P < 0.001 versus HCC827. GAPDH, glyceraldehyde-3-phosphate dehydrogenase. (C) Decay-corrected whole-body coronal PET images of HCC827, H1975, H520, and H358 tumor-bearing mice at 30, 60, and 120 min after injection of 3.7 megabecquerels (MBq; 100 μCi) of 18F-MPG (red arrows indicate the location of the tumors). % ID/g, percentage of injected dose per gram. (D) Comparison of tumor/muscle ratios of 18F-MPG and 18F-FDG at 60 min after injection of 3.7 MBq (100 μCi) of tracer in different tumor-bearing mice (n = 6 per group) as measured by PET imaging. Data are means ± SD. ***P < 0.001 versus line-matched 18F-MPG uptake in HCC827 cells, analysis of variance (ANOVA) with the Newman-Keuls multiple comparison test. 18F-FDG uptake versus 18F-MPG uptake in HCC827, ANOVA with the Newman-Keuls multiple comparison test.

  • Fig. 3 Study design and patient allocation.

    Seventy-five of 102 NSCLC patients were divided into four groups according to EGFR mutation status by gene test result at baseline interview. All patients had participated in the 18F-MPG study, and 73 of 75 patients participated in both the 18F-MPG and 18F-FDG studies. A total of 71 patients received EGFR-TKIs and/or continued to be treated with EGFR-TKIs.

  • Fig. 4 Representative CT and PET/CT images of three patients with NSCLCs.

    (A) Patient (number 3) with an EGFR exon 19 E746-A750 deletion who did not receive EGFR-TKI treatment, with a tumor in the upper right lobe of the lung [red arrows; 18F-MPG SUVmax (maximum standard uptake value) of 3.93]. (B) Patient (number 36) with an EGFR exon 21 L858R point mutation, 15 days after gefitinib treatment. Tumor in the upper right lung lobe (red arrows; 18F-MPG SUVmax of 3.60) and spinal metastasis (yellow arrows; 18F-MPG SUVmax of 2.11) were observed both on 18F-MPG and 18F-FDG images. (C) Patient (number 54) with wild-type EGFR who did not receive EGFR-TKI treatment. Tumor in the left hilum (red arrows; 18F-MPG SUVmax of 1.38) was prominently observed on the 18F-MPG image.

  • Fig. 5 EGFR mutation status of patients with NSCLC, detected by histological examination and ARMS PCR.

    (A) Histological and amplification refractory mutation system (ARMS) polymerase chain reaction (PCR) confirmation of a patient (number 3) with EGFR exon 19 deletion: hematoxylin and eosin (H&E) staining at ×10 magnification. Scale bar, 100 μm. Immunohistochemistry for total-EGFR, phosphorylated EGFR (phospho-EGFR), and E746-A750del–specific EGFR at ×20 magnification from inset in (B). Scale bars, 50 μm. ARMS PCR: ΔCt = 18.57 < 26 (cutoff). (B) Histological and ARMS PCR confirmation of a patient (number 36) with EGFR exon 21 L858R point mutation: H&E staining at ×10 magnification. Scale bar, 100 μm. Immunohistochemistry for total-EGFR, phospho-EGFR, and L858R-specific EGFR at ×20 magnification from inset in (B). Scale bars, 50 μm. ARMS PCR: ΔCt = 21.83 < 26 (cutoff). (C) Histological and ARMS PCR confirmation of a patient (number 54) with wild-type EGFR: H&E staining at ×10 magnification. Scale bar, 100 μm. Immunohistochemistry for total-EGFR, phospho-EGFR, and E746-A750 del–specific EGFR at ×20 magnification from inset in (C). Scale bars, 50 μm. ARMS PCR: ΔCt = 113 > 26 (cutoff).

  • Fig. 6 Correlation between 18F-MPG uptake and EGFR mutation status in patients with NSCLC and progression-free survival.

    (A and B) Box plot of 18F-MPG SUVmax (A) and 18F-FDG SUVmax (B) for patients with NSCLC with EGFR-activating mutation and wild-type EGFR, pre-TKI treatment and post-TKI treatment. Data are means ± SD. **P < 0.01 and ****P < 0.0001, ANOVA with the Newman-Keuls multiple comparison test. (C) Box plot of 18F-MPG SUVmax and 18F-FDG SUVmax for patients with NSCLC with EGFR exon 19 deletion and EGFR exon 21 L858R point mutation. No association was found between 18F-MPG uptake and EGFR exon 19 deletion or EGFR 21 point mutation. P = 0.8486, one-way ANOVA. (D) Receiver operating characteristic curve representing the sensitivity and specificity of 18F-MPG SUVmax for predicting the presence of EGFR mutation status in patients with NSCLC [area under the curve (AUC) = 0.8440, P < 0.0001]. (E) Kaplan-Meier plots of progression-free survival according to EGFR mutation. HR (95% CI), 0.2135 (0.04517 to 0.2925); P < 0.0001. (F) Kaplan-Meier plots of progression-free survival according to 18F-MPG SUVmax. HR (95% CI), 0.2083 (0.03406 to 0.2438); P < 0.0001.

  • Fig. 7 Examples of 18F-MPG PET/CT imaging monitoring the change of EGFR mutation status and EGFR-TKI therapy response in patients with NSCLC.

    (A) Patient with NSCLC with a shift in tumor from EGFR-activating mutation to EGFR wild type. Scans from a 47-year-old Asian male (number 9) with smoking history and lung adenocarcinoma are shown. Row 1: 18F-MPG PET/CT tumor SUVmax is 2.31 (red arrow), and EGFR gene test shows EGFR exon 19 E746-A750 deletion at 6 months after gefitinib treatment. Row 2: 18F-MPG PET/CT tumor SUVmax decreased to 2.01 (red arrow), and EGFR gene test with second fine-needle aspiration shows EGFR wild type at 20 months after gefitinib treatment. Row 3: 20 months after TKI 18F-MPG PET/CT also showed that a new tumor appeared in the inferior lobe (orange arrow) indicating progressive disease. H, heart; L, liver. (B) From a patient with NSCLC with unknown EGFR mutation status, an example of 18F-MPG PET/CT response: Scans from a 60-year-old Asian female (number 22) with no smoking history and lung adenocarcinoma are shown. Baseline 18F-MPG PET/CT tumor SUVmax is 3.04, and tumor size is 3.1 × 2.6 cm2 (red arrow). 18F-MPG PET/CT tumor SUVmax has decreased to 2.61 at 50 days after gefitinib treatment, and the tumor was shrunk [tumor size, 2.3 × 1.3 cm2 (red arrow)]. Marked 18F-MPG response was seen on day 50, and PET/CT scan showed partial response. K, kidney.

  • Table 1 Patient characteristics.

    CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease; WT, wild type; AJCC, American Joint Committee on Cancer.

    CharacteristicsTotalEGFR exon 19
    deletions
    EGFR exon 21 point
    mutations
    EGFR (WT)Unknown
    Patients, n (%)7517 (23%)20 (27%)33 (44%)5 (6%)
    Ages (years)
      Median6060606462
      Range45–8047–8045–7845–7845–80
    Sex (male/female)34:417:105:1518:154:1
    Smoking status
      Ever smoker3389160
      Never smoker42911175
    Stage (AJCC)
      I107120
      II31020
      III3059133
      IV32410162
    Histology
      Adenocarcinoma661720245
      Squamous cell carcinoma90090
    18F-MPG PET/CT
      SUVmax ≥2.23
      SUVmax <2.23
    42
    33
    16
    1
    16
    4
    6
    27
    4
    1
    TKI therapy response
    CR20200
    PR31111163
    SD112270
    PD2723202

Supplementary Materials

  • www.sciencetranslationalmedicine.org/cgi/content/full/10/431/eaan8840/DC1

    Materials and Methods

    Fig. S1. Synthetic scheme of T-MPG (compound 9) and F-MPG (compound 13).

    Fig. S2. Radioactivity chromatograms of 18F-MPG.

    Fig. S3. Characterization of HCC827, H1975, H520, and H358 cell lines and xenograft tumor models in vitro and ex vivo.

    Fig. S4. Predicted binding modes of F-MPG and PD153035 with the EGFR wild type, EGFR exon 19 E746-A750 deletion homology model, and EGFR L858R/T790M double mutant.

    Fig. S5. Uptake of MPG analogs in NSCLC cells.

    Fig. S6. Accumulation of 18F-MPG in different mice tumor xenografts.

    Fig. S7. Biodistribution of 18F-MPG and 18F-FDG in all patients with NSCLC at 60 min after injection.

    Fig. S8. Gene sequencing confirmation of an NSCLC patient (number 9) with a shift in tumor from EGFR-activating mutation to EGFR wild type.

    Table S1. Quantitative 18F-MPG PET ROI analysis of organ uptake in different tumor xenografts in mice at 30 min (n = 6 per group).

    Table S2. Quantitative 18F-MPG PET ROI analysis of organ uptake in different tumor xenografts in mice at 1 hour (n = 6 per group).

    Table S3. Quantitative 18F-MPG PET ROI analysis of organ uptake in different tumor xenografts in mice at 2 hours (n = 6 per group).

    Table S4. Quantitative 18F-FDG PET ROI analysis of organ uptake in different tumor xenografts in mice at 1 hour (n = 6 per group).

    Table S5. Biodistribution of 18F-MPG in different tumor xenografts in mice at 30 min (n = 6 per group).

    Table S6. Biodistribution of 18F-MPG in different tumor xenografts in mice at 1 hour (n = 6 per group).

    Table S7. Biodistribution of 18F-MPG in different tumor xenografts in mice at 2 hours (n = 6 per group).

    Table S8. First-in-human data, OD in microsievert/megabecquerel.

    Table S9. 18F-MPG and 18F-FDG SUVmax in all NSCLC patient organs 60 min after injection.

    Table S10. Tumor accumulation of 18F-MPG and 18F-FDG in different groups of NSCLC patients at 60 min.

  • Supplementary Material for:

    A PET imaging approach for determining EGFR mutation status for improved lung cancer patient management

    Xilin Sun, Zunyu Xiao, Gongyan Chen, Zhaoguo Han, Yang Liu, Chongqing Zhang, Yingying Sun, Yan Song, Kai Wang, Fang Fang, Xiance Wang, Yanhong Lin, Lili Xu, Liming Shao, Jin Li, Zhen Cheng,* Sanjiv Sam Gambhir,* Baozhong Shen*

    *Corresponding author. Email: shenbz{at}ems.hrbmu.edu.cn (B.S.); sgambhir{at}stanford.edu (S.S.G.); zcheng{at}stanford.edu (Z.C.)

    Published 7 March 2018, Sci. Transl. Med. 10, eaan8840 (2018)
    DOI: 10.1126/scitranslmed.aan8840

    This PDF file includes:

    • Materials and Methods
    • Fig. S1. Synthetic scheme of T-MPG (compound 9) and F-MPG (compound 13).
    • Fig. S2. Radioactivity chromatograms of 18F-MPG.
    • Fig. S3. Characterization of HCC827, H1975, H520, and H358 cell lines and xenograft tumor models in vitro and ex vivo.
    • Fig. S4. Predicted binding modes of F-MPG and PD153035 with the EGFR wild type, EGFR exon 19 E746-A750 deletion homology model, and EGFR L858R/T790M double mutant.
    • Fig. S5. Uptake of MPG analogs in NSCLC cells.
    • Fig. S6. Accumulation of 18F-MPG in different mice tumor xenografts.
    • Fig. S7. Biodistribution of 18F-MPG and 18F-FDG in all patients with NSCLC at 60 min after injection.
    • Fig. S8. Gene sequencing confirmation of an NSCLC patient (number 9) with a shift in tumor from EGFR-activating mutation to EGFR wild type.
    • Table S1. Quantitative 18F-MPG PET ROI analysis of organ uptake in different tumor xenografts in mice at 30 min (n = 6 per group).
    • Table S2. Quantitative 18F-MPG PET ROI analysis of organ uptake in different tumor xenografts in mice at 1 hour (n = 6 per group).
    • Table S3. Quantitative 18F-MPG PET ROI analysis of organ uptake in different tumor xenografts in mice at 2 hours (n = 6 per group).
    • Table S4. Quantitative 18F-FDG PET ROI analysis of organ uptake in different tumor xenografts in mice at 1 hour (n = 6 per group).
    • Table S5. Biodistribution of 18F-MPG in different tumor xenografts in mice at 30 min (n = 6 per group).
    • Table S6. Biodistribution of 18F-MPG in different tumor xenografts in mice at 1 hour (n = 6 per group).
    • Table S7. Biodistribution of 18F-MPG in different tumor xenografts in mice at 2 hours (n = 6 per group).
    • Table S8. First-in-human data, OD in microsievert/megabecquerel.
    • Table S9. 18F-MPG and 18F-FDG SUVmax in all NSCLC patient organs 60 min after injection.
    • Table S10. Tumor accumulation of 18F-MPG and 18F-FDG in different groups of NSCLC patients at 60 min.

    [Download PDF]

Navigate This Article