Jingjing Fu1, Yan Xie1, Tong Fu2, Fan Qiu1, Fei Yu1, Wei Qu1, Xiaochen Yao1, Aiping Zhang3, Zhenhua Yang4, Guoqiang Shao1, Qingle Meng1, Xiumin Shi1, Yue Huang5, Wei Gu6, Feng Wang7. 1. Department of Nuclear Medicine, Nanjing First Hospital, Nanjing Medical University, 68 Changle Road, Nanjing, 210006, China. 2. Department of Imaging, Nanjing First Hospital, Nanjing Medical University, Nanjing, 210006, China. 3. Department of Thoracic Surgery, Nanjing First Hospital, Nanjing Medical University, Nanjing, 210006, China. 4. Department of Respiratory, Nanjing First Hospital, Nanjing Medical University, 68 Changle Road, Nanjing, 210006, China. 5. Department of Pathology, Nanjing First Hospital, Nanjing Medical University, 68 Changle Road, Nanjing, 210006, China. huangyuexxy@163.com. 6. Department of Respiratory, Nanjing First Hospital, Nanjing Medical University, 68 Changle Road, Nanjing, 210006, China. guw2001@126.com. 7. Department of Nuclear Medicine, Nanjing First Hospital, Nanjing Medical University, 68 Changle Road, Nanjing, 210006, China. fengwangcn@hotmail.com.
Abstract
BACKGROUND: Epidermal growth factor receptor tyrosine kinase inhibitors (TKIs) are beneficial in patients with lung cancer. We explored the clinical value of [99mTc]Tc-Galacto-RGD2 single-photon emission computed tomography (SPECT/CT) in patients with lung cancer, integrin αvβ3 expression, and neovascularization in lung cancer subtypes was also addressed. METHODS: A total of 185 patients with lung cancer and 25 patients with benign lung diseases were enrolled in this prospective study from January 2013 to December 2016. All patients underwent [99mTc]Tc-Galacto-RGD2 imaging. The region of interest was drawn around each primary lesion, and tumour uptake of [99mTc]Tc-Galacto-RGD2 was expressed as the tumour/normal tissue ratio(T/N). The diagnostic efficacy was evaluated by receiver operating characteristic curve analysis. Tumour specimens were obtained from 66 patients with malignant diseases and 7 with benign disease. Tumour expression levels of αvβ3, CD31, Ki-67, and CXCR4 were further analysed for the evaluation of biological behaviours. RESULTS: The lung cancer patients included 22 cases of small cell lung cancer (SCLC), 48 squamous cell carcinoma (LSC), 97 adenocarcinoma (LAC), and 18 other types of lung cancer. The sensitivity, specificity, and accuracy of [99mTc]Tc-Galacto-RGD2 SPECT/CT using a cut-off value of T/N ratio at 2.5 were 91.89%, 48.0%, and 86.67%, respectively. Integrin αvβ3 expression was higher in non-SCLC compared with SCLC, while LSC showed denser neovascularization and higher integrin αvβ3 expression. Integrin αvβ3 expression levels were significantly higher in advanced (III, IV) than early stages (I, II). However, there was no significant correlation between tumour uptake and αvβ3 expression. CONCLUSIONS: [99mTc]Tc-Galacto-RGD2 SPECT/CT has high sensitivity but limited specificity for detecting primary lung cancer, integrin expression in the tumour vessel and tumour cell membrane contributes to the tumour uptake.
BACKGROUND:Epidermal growth factor receptortyrosine kinase inhibitors (TKIs) are beneficial in patients with lung cancer. We explored the clinical value of [99mTc]Tc-Galacto-RGD2 single-photon emission computed tomography (SPECT/CT) in patients with lung cancer, integrin αvβ3 expression, and neovascularization in lung cancer subtypes was also addressed. METHODS: A total of 185 patients with lung cancer and 25 patients with benign lung diseases were enrolled in this prospective study from January 2013 to December 2016. All patients underwent [99mTc]Tc-Galacto-RGD2 imaging. The region of interest was drawn around each primary lesion, and tumour uptake of [99mTc]Tc-Galacto-RGD2 was expressed as the tumour/normal tissue ratio(T/N). The diagnostic efficacy was evaluated by receiver operating characteristic curve analysis. Tumour specimens were obtained from 66 patients with malignant diseases and 7 with benign disease. Tumour expression levels of αvβ3, CD31, Ki-67, and CXCR4 were further analysed for the evaluation of biological behaviours. RESULTS: The lung cancerpatients included 22 cases of small cell lung cancer (SCLC), 48 squamous cell carcinoma (LSC), 97 adenocarcinoma (LAC), and 18 other types of lung cancer. The sensitivity, specificity, and accuracy of [99mTc]Tc-Galacto-RGD2 SPECT/CT using a cut-off value of T/N ratio at 2.5 were 91.89%, 48.0%, and 86.67%, respectively. Integrin αvβ3 expression was higher in non-SCLC compared with SCLC, while LSC showed denser neovascularization and higher integrin αvβ3 expression. Integrin αvβ3 expression levels were significantly higher in advanced (III, IV) than early stages (I, II). However, there was no significant correlation between tumour uptake and αvβ3 expression. CONCLUSIONS: [99mTc]Tc-Galacto-RGD2 SPECT/CT has high sensitivity but limited specificity for detecting primary lung cancer, integrin expression in the tumour vessel and tumour cell membrane contributes to the tumour uptake.
Lung cancer is the leading cause of cancermortality worldwide [1, 2]. The incidence and mortality of lung cancer in China have increased rapidly in the last three decades, associated with increases in air pollution and tobacco consumption [3, 4]. However, new clinical treatment strategies, such as antiangiogenic epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs) and immunotherapy, have significantly improved the outcomes of patients with lung cancer in the last decade [5]. TKIs have a cytostatic effect on tumour cells by slowing their growth and preventing the development of distant metastases [6, 7]. Multiplex genetic sequencing has been used to select appropriate treatment, based on the recommendation of the American Society of Clinical Oncology (ASCO); however, this requires obtaining enough tumour tissue by biopsy or surgery. Unfortunately, suitable tumour specimens are unavailable for some patients due to the tumour heterogeneity or undetermined primary lesion.Angiogenesis plays important roles in tumour initiation, development, and metastasis [8]. Integrins are a diverse family of glycoproteins that form heterodimeric receptors for extracellular matrix molecules [9-11], of which integrin αvβ3, with an exposed arginine-glycine-aspartate (RGD) tripeptide sequence, is the most-extensively studied [11]. Integrin αvβ3 is highly expressed in the neovasculature in solid tumours, including neuroblastoma, osteosarcoma, glioblastoma, breast cancer, and prostate cancer [12-20]. The highly restricted expression of integrin αvβ3 in normal tissues compared with its overexpression in tumour cells suggests that it may provide an interesting molecular target for the early detection of malignant tumours [12]. Overexpression of integrin αvβ3 was also correlated with tumour invasiveness in breast cancer, indicating a possible role in evaluating metastatic potential [19].Radiolabelled RGD peptide as a target ligand for angiogenesis imaging has been well documented in preclinical and clinical studies [12, 21, 22]. In a previous multicentre study, we showed that [99mTc]Tc labelled RGD dimers, such as [99mTc]Tc-3PRGD2, had high sensitivity for the detection of lung cancer, including primary and metastatic tumours [21, 23, 24]. [99mTc]Tc-Galacto-RGD2, with higher affinity to αvβ3 and a favourable biodistribution, has been synthesized and utilized for the quantitative evaluation of αvβ3 expression and of tumour angiogenesis [25].Clinically, multiple lymphadenopathy and remote metastasis were developed rapidly in higher aggressive lung cancer even with radical resection and comprehensive treatment, we suppose some key molecules medicate the tumour development and metastasis. Therefore, we conducted a longitudinal study to evaluate the clinical role of [99mTc]Tc-Galacto-RGD2 SPECT/CT in a large population of patients with lung neoplasms. We also explored the expression of integrin αvβ3 protein in tumour cells and in the neovasculature, and determined the capability of the technique to detect lymphadenopathy and bone metastasis in patients with advanced lung cancer. Herein, we investigated the value of RGD-based imaging as a surrogate for molecular phenotyping in lung cancer.
Methods
Patients
This prospective, single-centre study enrolled patients referred to our centre with suspected lung neoplasms from January 2013 to December 2016. [99mTc]Tc-Galacto-RGD2 SPECT/CT was performed in all patients; the final diagnosis was confirmed by histopathology based on acupuncture biopsy or surgery. A total of 210 consecutive patients (147 male, 63 female; mean age 63.80 ± 10.51 years, range 21–85 years) were enrolled and analysed. Of the 210 patients, 185 were confirmed with lung cancer and the other 25 patients had benign pulmonary diseases and served as the control. Patients who had undergone perioperative chemotherapy or radiotherapy were excluded from this study; the schema of study is shown in Fig. 1.
Fig. 1
The study flow for the [99mTc]Tc-Galacto-RGD2 imaging in the diagnosis of lung cancer
The study flow for the [99mTc]Tc-Galacto-RGD2 imaging in the diagnosis of lung cancer
[99mTc]Tc-Galacto-RGD2 radiolabelling and quality control
[99mTc]Tc-Galacto-RGD2 labelling was carried out as described previously [25]. The Galacto-RGD2 was friendly offered by the School of Health Sciences, Purdue University (Indiana, USA). Chemicals were purchased from Sigma-Aldrich (St. Louis, MO). Na[99mTc]TcO4 was obtained from DongCheng Pharmaceutical Company (Nanjing, China). Briefly, radiolabelling with performed with a lyophilized kit formulation containing 20 μg, 7 mg TPPTS (trisodium triphenylphosphine-3,3′,3″-trisulfonate), 6.5 mg tricine, 40 mg mannitol, 38.5 mg disodium succinate hexahydrate, and 12.7 mg succinic acid. 99mTc-labelling was accomplished by adding 1–1.5 mL of Na[99mTc]TcO4 solution (1,110–1,850 MBq). The reconstituted vial was heated at 100 °C for 30 min and the resulting solution was analysed by radio-high-performance liquid chromatography using a Lab Alliance system equipped with a ram IN-US detector and Zorbax C18 column (4.6 mm × 250 mm, 300 Å pore size, Waters Xbridge C18, Milford, MA). The flow rate was 1 mL/min, the mobile phase was isocratic with 90% solvent A (25 mM NH4OAc buffer, pH 6.8) and 10% solvent B (acetonitrile) at 0–5 min, followed by a gradient mobile phase from 10% B at 5 min to 40% B at 20 min. The radiochemical purity was > 95% for all imaging.
[99mTc]Tc-Galacto-RGD2 imaging and interpretation
The radiochemical purity was 95.1% ± 2.9%. [99mTc]Tc-Galacto-RGD2 was administered at 555–740 MBq (15–20 mCi) and whole-body images were acquired at 1-h post-injection. The chest images, including the upper abdomen and adrenal glands, were performed using a combined transmission and emission device with x-ray tube and detector. All-purpose collimator centred on the140-keV energy peak with a 20% symmetrical energy window. Thirty projection images were acquired over a 180° arc at 6° intervals for each SPECT head. The acquisition time was 30 s at each projection. The transaxial data were reconstructed using Ordered Subset Expectation Maximization, 2 iterations, 8 subsets (Symbia T6 SPECT/CT; Siemens AG, Germany). Anatomic CT images were performed for attenuation correction and tumour localization. If unexpected lesions were detected by whole-body imaging, additional abdomen or pelvis images were also acquired.All images were interpreted independently on the computer monitor in three orthogonal planes by nuclear medicine physicians and a radiologist who were unaware of the clinical information and other imaging examinations. Significantly greater local uptake of [99mTc]Tc-Galacto-RGD2 compared with the adjacent surrounding lung was interpreted as demonstrating a malignant lesion, and uptake less than or equal to the adjacent or surrounding lung was interpreted as a benign lesion. Focal activity in the hilum and mediastinum greater than the surrounding mediastinal activity was interpreted as lymphadenopathy. Regions of interest (ROI) were drawn around the primary lesion and contralateral lung tissue, respectively, and [99mTc]Tc-Galacto-RGD2 uptake was measured and expressed as the tumour/normal tissue ratio (T/N).
Composite reference standard
All available cytologic, histologic, follow-up, and imaging findings were used as a composite reference standard for the presence of tumour lesions. This is considered the optimal gold standard because cytologic or histologic verification of every lesion was not feasible or justifiable in these patients. Whenever possible, new findings on [99mTc]Tc-Galacto-RGD2 SPECT-CT were verified by additional investigations.
Immunohistochemistry (IHC) analysis
Tumour specimens were obtained from patients who underwent complete resection or biopsy. The sections were fixed in formalin, embedded in paraffin, deparaffinized, and stained with haematoxylin and eosin (H&E). Integrin αvβ3, Ki-67, CXCR4, and CD31 expression were analysed by IHC to evaluate the biological tumour behaviour. Sections were cut at 3-μm, dewaxed in xylene, and rehydrated in graded ethanols. Integrin αvβ3 and CXCR4 expression, microvessel density (CD31), and tumour cell proliferation (Ki-67) were detected by incubating the slides with monoclonal antibodies against human integrin αvβ3 (1:200, sc-7312; Santa Cruz Biotechnology, Santa Cruz, California, US), CXCR4 (1:100, ab227767; Abcam, Massachusetts, US), Ki-67 (1:100, ab270650; Abcam), or CD31 (1:50, ab28364; Abcam), respectively, overnight, followed by horseradish peroxidase-conjugated anti-mouse IgG (1:1000, Earth Ox, Millbrae, California, US) with 3′3-diaminobenzidine as the chromogen. H&E staining was also performed. All images were obtained at 100 × magnification with the same exposure time. Brightness and contrast were adjusted similarly in all images. Integrin αvβ3 and CXCR4 expression levels were quantified by determining the optical density (OD) after immunostaining.
Statistical analysis
All statistical analyses were carried out using R software (version 3.6.1) and graphs were constructed using GraphPad Prism software (version 7). Continuous variables with a non-normal distribution were expressed as median (interquartile range). Differences in T/NT and protein expression levels among groups were compared using Wilcoxon’s rank-sum or Kruskal–Wallis tests. The sensitivity, specificity, area under the curve (AUC), and cut-off value of T/NT were evaluated by receiver operating characteristic curve (ROC) analysis. Correlations between continuous variables with non-normal distributions were evaluated by Spearman’s rank correlation analysis. Bonferroni’s correction was applied for multiple comparisons. Statistical significance was established at p < 0.05.
Results
Patient characteristics
The clinical characteristics of the patients are shown in Table 1. Of the 210 consecutive patients enrolled in this study, 185 (88.1%) had malignant neoplasms identified by histopathology, including 22 patients with small cell lung cancer (SCLC), 97 with adenocarcinoma (LAC), 48 with squamous cell carcinoma (LSC), and 18 patients with other malignant lung tumours. Tumour tissues were obtained during thoracic surgery (n = 118), fine-needle aspiration (n = 35), or bronchoscopy (n = 32). Of the 25 patients with benign respiratory diseases, the benign nature of the lesion was confirmed during clinical follow-up in 12 patients, by histopathology in 7 patients, and at imaging follow-up in 6 patients. According to the Tumour, Node, and Metastasis (TNM) classification of lung cancer 8th edition published in 2015 [26], 37 patients were diagnosed with stage I (20.00%), 13 with stage II (7.03%), 40 with stage III (21.62%), and 95 patients with stage IV (51.35%). The volume of the primary tumour (median (interquartile range): 28.01 (12.30, 76.33) mm3 was significantly higher in patients with malignant compared with benign disease (10.89 (8.66, 15.77) mm3) (Wilcoxon’s rank-sum test, p < 0.01).
Table 1
Clinical characteristics of 210 subjects
Variants
Lung cancer
Benign disease
p
General
Age (years)
64.17 ± 10.15
61.04 ± 12.77
0.25
Sex
Male
133 (56.00%)
14 (71.89%)
Female
52 (44.00%)
11 (28.11%)
0.16
Cancer type
LAC
97 (52.43%)
LSC
48 (25.95%)
SCLC
22 (11.89%)
Other
18 (9.73%)
Stage
I
37 (20.00%)
II
13 (7.03%)
III
40 (21.62%)
IV
95 (51.35%)
LAC adenocarcinoma, LSC squamous cell carcinoma, SCLC small cell lung cancer
Clinical characteristics of 210 subjectsLACadenocarcinoma, LSC squamous cell carcinoma, SCLC small cell lung cancerHigh-contrast images acquired 1 h after injection of [99mTc]Tc-Galacto-RGD2 showed higher focal uptake in malignant primary tumours and metastatic lymph nodes (Fig. 2), compared with significantly lower uptake in benign lesions, the ratio (median (interquartile range)) of T/N in malignant disease was 6.84 (4.62, 9.86), whereas that of benign diseases was 2.53 (1.24, 3.91), p < 0.01. We also compared the uptake in different lung cancer subtypes (Fig. 3). [99mTc]Tc-Galacto-RGD2 uptake was highest in LSC (T/N: 8.53 (6.75, 10.99)), followed by LAC (T/N: 6.84 (4.64, 9.07)) and SCLC (T/NT: 4.73 (2.47, 5.85)). Other types of lung cancer (T/NT: 5.23 (3.32, 11.50)) showed moderate uptake of [99mTc]Tc-Galacto-RGD2 in the primary tumour, with no significant difference between other types and LSC, LAC, and SCLC. We also compared uptake by the primary tumour between locoregional and advanced stages. T/N was significantly lower in stage I–II (5.78 (3.62, 7.95)) compared with advanced stages (III –IV; 7.28 (5.43, 10.34)), p < 0.01. However, there was overlap with inflammatory pseudotumours or tuberculosis. RGD avidity was found in in two cases of pulmonary sequestration and thymoma, respectively, due to higher density of micro-vessels (Figs. 4, 5). ROC analysis indicated that the sensitivity, specificity, and accuracy of [99mTc]Tc-Galacto-RGD2 were 91.89%, 48.0%, and 86.67%, respectively, using a cut-off value of 2.5. With a T/N cut-off value of 3.94, the AUC was 0.83 and the sensitivity and the specificity were 82.7% and 76.0%, respectively.
Fig. 2
[99mTc]Tc-Galacto-RGD2 imaging showed RGD-avid uptake in the primary tumour, lymphadenopathy, and remote metastases in a patient with suspected multiple myeloma. The final diagnosis was lung adenocarcinoma confirmed by bronchoscopic biopsy. a Primary lesion presented in the lung window. b, c Enhanced CT showed primary tumour and lymphadenopathy in the right hilum. d, e Bone scan showed lytic lesion in the right rib and sclerotic lesions in the pelvis. f, g
99mTc-Galacto-RGD2 image showed avid lesions in the right lung and the right hilum. h, i, j IHC staining. h αvβ3 expression. i CD31 expression in neo-vasculature. j CXCR4 expression in tumour tissue
Fig. 3
Ratio of primary tumour to normal tissue (T/NT) in different groups. *p < 0.05; **p < 0.01; ***p < 0.001
Fig. 4
[99mTc]Tc-Galacto-RGD2 imaging showed focal uptake in a patient with pulmonary sequestration. a, b Enhanced CT showed the tumour blood supply originated from the abdominal aorta. c, d RGD imaging showed focal uptake in the tissue, and necrosis in the central zone of the tissue
Fig. 5
Female patient with an anterior mediastinum lesion, the final diagnosis was thymus adenoma. a–d [99mTc]Tc-Galacto-RGD2 imaging showed high focal uptake. e The surface of the tumour showed many new vessels. f, g IHC showed a high density of neovascularization with high expression of CD31, but no significant αvβ3 expression in the tumour cell membrane. h H&E staining
[99mTc]Tc-Galacto-RGD2 imaging showed RGD-avid uptake in the primary tumour, lymphadenopathy, and remote metastases in a patient with suspected multiple myeloma. The final diagnosis was lung adenocarcinoma confirmed by bronchoscopic biopsy. a Primary lesion presented in the lung window. b, c Enhanced CT showed primary tumour and lymphadenopathy in the right hilum. d, e Bone scan showed lytic lesion in the right rib and sclerotic lesions in the pelvis. f, g
99mTc-Galacto-RGD2 image showed avid lesions in the right lung and the right hilum. h, i, j IHC staining. h αvβ3 expression. i CD31 expression in neo-vasculature. j CXCR4 expression in tumour tissueRatio of primary tumour to normal tissue (T/NT) in different groups. *p < 0.05; **p < 0.01; ***p < 0.001[99mTc]Tc-Galacto-RGD2 imaging showed focal uptake in a patient with pulmonary sequestration. a, b Enhanced CT showed the tumour blood supply originated from the abdominal aorta. c, d RGD imaging showed focal uptake in the tissue, and necrosis in the central zone of the tissueFemale patient with an anterior mediastinum lesion, the final diagnosis was thymus adenoma. a–d [99mTc]Tc-Galacto-RGD2 imaging showed high focal uptake. e The surface of the tumour showed many new vessels. f, g IHC showed a high density of neovascularization with high expression of CD31, but no significant αvβ3 expression in the tumour cell membrane. h H&E staining
Histopathology and IHC
Of the 210 patients with suspected lung cancer, immunohistochemistry (IHC) was performed in 66 patients with lung cancer and seven patients with benign diseases. Expression levels (median (interquartile range)) of integrin αvβ3 were significantly higher in lung cancer (OD: 15,020.5 (4482.6, 44,455.2)) compared with benign diseases (OD: 1797.8 (794.0, 2943.6); p < 0.01) (Table 2). CD31 levels were also elevated in lung cancer (OD: 21.9 (13.75, 34.35) vs 9.00 (8.90, 11.50); p < 0.01). Higher levels of integrin αvβ3 were expressed in advanced tumours (OD: 19,729.00 (6445.40, 45,288.30)) compared with locoregional tumours (5914.40 (1461.60, 17,658.20)), p < 0.05. Integrin αvβ3 was highly expressed not only in endothelial cells in the neovasculature, reflected by CD31 expression, but also in tumour cells (Fig. 6), with a higher density of neovasculature and integrin αvβ3 expression in the primary tumour. Integrin αvβ3 was also significantly correlated with CD31 expression in lung cancer (r = 0.30, p = 0.016). However, there was no correlation between tumour uptake of [99mTc]Tc-Galacto-RGD2 and integrin αvβ3 expression in the primary tumour in this study (Fig. 7). Squamous lung cancer usually showed higher level of αvβ3 in the tumour cell and the higher density of microvessel, which was consistent with RGD imaging as shown in Fig. 8. Aggressive LAC tends to higher express integrin αvβ3 in the tumour cell and has denser micro-vessels, which showed focal uptake in the RGD image, as shown in Fig. 9. Neo-vascularization varied in benign respiratory diseases, associated with higher integrin αvβ3 expression. In the current study, integrin αvβ3 correlated with CD31 expression in the neo-vessel, indicating that integrin αvβ3 mediated angiogenesis, leading to tumour development and metastasis. We also examined CXCR4 expression. CXCR4 was highly expressed in lung cancer, as demonstrated by IHC. Furthermore, expression levels of CXCR4 tended to be positively correlated with integrin αvβ3 levels in lung cancer specimens (r = 0.22, p > 0.05). In addition, the proliferation index (Ki-67, median (interquartile range)) in LSC and SCLC (27.45 (11.88, 42.00) and 70.00 (55.13, 73.48), respectively) were both significantly higher than in LAC (10.15 (2.98, 27.89)) (Table 3).
Table 2
Immunohistochemical results of lung cancer and benign disease
Variants
Lung cancer (n = 66)
Benign disease (n = 7)
p
Integrin αvβ3 (OD)
15,020.5 (4482.6,44,455.2)
1797.8 (794.0,2943.6)
1.08E−03
CXCR4 (OD)
5120.0 (1978.0,18,460.0)
538.6 (300.0,7101.7)
0.08
CD31 (MVD)
21.9 (13.75,34.35)
9.00 (8.90,11.50)
5.56E-03
Ki-67 (%)
20.00 (7.46,40.00)
Bold values indicate p values means there is a significance of the statistical results
OD optical density, MVD microvessel density. The non-normal distribution data showed as median (interquartile range)
Fig. 6
[99mTc]Tc-Galacto-RGD2 SPECT/CT detected the primary tumour and multiple lymph node metastases in advanced adenocarcinoma. [99mTc]Tc-Galacto-RGD2 image (a–d). a Whole-body image. b–d SPECT/CT. e–g IHC showed higher expression of αvβ3 in the tumour cells and neo-vasculature, higher density of micro-vessel with CD31 expression in the tumour tissue, and a higher Ki-67 index
Fig. 7
The correlation analysis between ratio of primary tumour to normal tissue (T/NT) and αvβ3 expression in different groups
Fig. 8
[99mTc]Tc-Galacto-RGD2 SPECT/CT showed RGD-avid lesion in a female patient with squamous cell carcinoma. a–d RGD imaging showed focal uptake in the left lung. e–g FDG-avid lesion shown in the left upper lung. h, i IHC showed a moderate αvβ3 expression in the tumour cell membrane, and a high density of neovascularization with high expression of CD31
Fig. 9
RGD-avid lesion in a patient with lung adenocarcinoma, consistent with αvβ3 expression which confirmed by IHC, flow cytometry and western blot. a–c [99mTc]Tc-Galacto-RGD2 image. d, e IHC showed higher expression of αvβ3 in the tumour cells and higher density of micro-vessel (CD31). f Flow cytometry showed higher expression of integrin β3 in tumour tissue compared with normal lung tissue. g Western blot showed higher integrin β3 expression in primary tumour tissue compared with normal lung tissue
Table 3
Immunohistochemical results among different groups of lung cancer
Stage
Class
T/NT (1 h)
I-II (n = 21)
III-IV (n = 45)
p
LAC (n = 34)
LSC (n = 26)
SCLC (n = 6)
p
low (< 7.21) (n = 33)
high (≥ 7.21) (n = 33)
p
Integrin αvβ3 (OD)
5914.4 (1461.6, 17,658.2)
19,729.0 (6445.4, 45,288.3)
3.10E-02
17,308.1 (7973.6,47,755.9)
9721.6 (3186.5,23,010.7)
6485.0 (3083.0,28,119.0)
0.31
16,958.0 (3842.8,36,479.0)
12,235.4 (4935.2,47,323.5)
0.98
CXCR4 (OD)
8438.0 (2582.5,22,821.8)
4002 (1863,11,541)
0.18
4536 (1348,10,501)
9052.0 (3688.9,33,442.9)
2324 (1980,4395)
0.11
6561.00 (2049.20,18,854.50)
5007 (1335,15,314)
0.65
CD31 (MVD)
28.20 (14.25,34.40)
21.33 (13.75,32.00)
0.54
27.80 (14.44,34.35)
19.72 (15.06,32.11)
13.32 (11.75,20.04)
0.25
25.80 (15.00,35.40)
18.20 (13.75,28.40)
0.25
Ki67 (%)
11.00 (5.00,40.00)
22.82 (8.63,43.95)
0.47
10.15 (2.98,27.89)
27.45 (11.88,42.00)
70.00 (55.13,73.48)
5.64E-03
16.13 (4.36,50.00)
21.90 (10.00,38.90)
0.55
Bold values indicate p values means there is a significance of the statistical results
LAC adenocarcinoma, LSC squamous cell carcinoma, SCLC small cell lung cancer, OD optical density, MVD microvessel density. The non-normal distribution data showed as median (interquartile range)
Immunohistochemical results of lung cancer and benign diseaseBold values indicate p values means there is a significance of the statistical resultsOD optical density, MVD microvessel density. The non-normal distribution data showed as median (interquartile range)[99mTc]Tc-Galacto-RGD2 SPECT/CT detected the primary tumour and multiple lymph node metastases in advanced adenocarcinoma. [99mTc]Tc-Galacto-RGD2 image (a–d). a Whole-body image. b–d SPECT/CT. e–g IHC showed higher expression of αvβ3 in the tumour cells and neo-vasculature, higher density of micro-vessel with CD31 expression in the tumour tissue, and a higher Ki-67 indexThe correlation analysis between ratio of primary tumour to normal tissue (T/NT) and αvβ3 expression in different groups[99mTc]Tc-Galacto-RGD2 SPECT/CT showed RGD-avid lesion in a female patient with squamous cell carcinoma. a–d RGD imaging showed focal uptake in the left lung. e–g FDG-avid lesion shown in the left upper lung. h, i IHC showed a moderate αvβ3 expression in the tumour cell membrane, and a high density of neovascularization with high expression of CD31RGD-avid lesion in a patient with lung adenocarcinoma, consistent with αvβ3 expression which confirmed by IHC, flow cytometry and western blot. a–c [99mTc]Tc-Galacto-RGD2 image. d, e IHC showed higher expression of αvβ3 in the tumour cells and higher density of micro-vessel (CD31). f Flow cytometry showed higher expression of integrin β3 in tumour tissue compared with normal lung tissue. g Western blot showed higher integrin β3 expression in primary tumour tissue compared with normal lung tissueImmunohistochemical results among different groups of lung cancerBold values indicate p values means there is a significance of the statistical resultsLACadenocarcinoma, LSC squamous cell carcinoma, SCLC small cell lung cancer, OD optical density, MVD microvessel density. The non-normal distribution data showed as median (interquartile range)
Lymphadenopathy and distant metastasis
Of the 185 patients with lung cancer, 116 patients had lymphadenopathy, 87 had remote metastasis, 17 had multiple lung tumours including pleural invasion, and 70 patients had bone metastasis. The metastatic lymph nodes and remote metastases showed high focal uptake of [99mTc]Tc-Galacto-RGD2. However, although lymphadenopathy was evaluated by imaging follow-up, the final diagnosis was not confirmed, and we were therefore unable to evaluate the diagnostic value of [99mTc]Tc-Galacto-RGD2 imaging for lymphadenopathy and remote metastasis in this study.
Discussion
We previously validated the ability of [99mTc]Tc-Galacto-RGD2 to identify iodine-refractory status in patients with thyroid cancer [27]. In a rare case with a solitary fibrous tumour located in the main pulmonary artery, [99mTc]Tc-Galacto-RGD2 imaging played an important role in detecting the primary tumour and predicting the metastatic potential [22]. In the current study, we evaluated the use of [99mTc]Tc-Galacto-RGD2 SPECT/CT for the detection of lung cancer. We also explored the expression of integrin αvβ3 and CXCR4 in different lung cancer subtypes, and compared the neovasculature among these subtypes. The correlations between tumour uptake of [99mTc]Tc-Galacto-RGD2 and integrin αvβ3 expression and neovascularization were also explored.High-contrast images of [99mTc]Tc-Galacto-RGD2 showed a significantly higher T/NT ratio in malignant compared with benign lung lesions. Malignant primary tumours and metastatic lymph nodes showed higher focal uptake, while benign lesions showed significantly lower uptake. [99mTc]Tc-Galacto-RGD2 SPECT/CT showed high sensitivity for primary tumours and remote metastases. ROC analysis showed a sensitivity and accuracy of 91.89% and 86.67%, respectively, for [99mTc]Tc-Galacto-RGD2 SPECT/CT, using a cut-off value of T/N ratio at 2.5. However, the specificity for differentiating between malignant and benign disease was limited, possibly because of the involvement of integrin αvβ3 in various benign diseases. Overlap usually occurs between tuberculosis and inflammatory pseudo-tumours, which usually show higher uptake of [99mTc]Tc-Galacto-RGD2 than other types of benign diseases, such as pneumonia [12].In the current study, IHC showed that αvβ3 levels were higher in advanced lung cancer, and proliferation index, represented by Ki-67, was significantly increased in advanced stages of SCLC, associated with metastatic potential [12, 19, 28]. Patients with lung cancer, even in the early stages, may develop multiple metastases several months even after thorough tumour resection, possibly related to specific tumour types with higher metastatic potential [19]. In the current study, CXCR4 expression levels were higher in lung cancer compared with benign disease, though the differences were not significant. Its expression was correlated with both integrin αvβ3 and CD31 expression in primary lung tumours, while integrin αvβ3 was also correlated with CD31. These findings validate our hypothesis that lymphadenopathy and remote metastasis are mediated by specific biological molecules. Integrin αvβ3 and CXCR4 may mediate angiogenesis, which may further promote lymph node and remote metastases. Integrin αvβ3-targeted imaging thus improves our understanding of the interactions between cancer cells and their microenvironment, which is a necessary prerequisite for the development of treatment strategies. This study showed higher levels of integrin αvβ3 were expressed in advanced tumours, integrin αvβ3 was also highly expressed not only in endothelial cells in the neo-vasculature but also in tumour cells, higher uptake was found in the primary tumour with a higher density of neo-vasculature and higher αvβ3 expression, which associated with multiple lymphadenopathy and remote metastasis. This finding confirmed integrin αvβ3 overexpression as an important component of tumour microenvironment, which was related with tumourigenic and aggressive behaviour in lung cancer. CXCR4 has been implicated in the chemotactic migration of cancer cells [10]. CXCR4 and integrin might synergistically promote lymphatic metastasis in lung cancer, and act as clinical predictors of lymph node metastasis in non-SCLC [29, 30]. High expression levels of chemokines are related to a poor prognosis and a poor chemotherapy tolerance in cancerpatients [31-34]. CXCR4 is a chemokine receptor that plays a critical role in the process of lymphocyte homing to lymphatic vessels and secondary lymphoid organs, including the lymph nodes [35].Integrin αvβ3 was expressed not only in the tumour cells, but also in the endothelium, though there was a lack of a correlation between tumour uptake of [99mTc]Tc-Galacto-RGD2 and integrin αvβ3 expression related to the heterogenicity of lung cancer. We found that tumour uptake of [99mTc]Tc-Galacto-RGD2 was related to integrin αvβ3 expression, neovascularization, and tumour stage, and integrin αvβ3 expression in tumour cells may promote lymphatic and distant metastases (Fig. 2). However, benign diseases showed variable degrees of angiogenesis, also associated with higher expression of integrin αvβ3, as shown in one patient with thymus adenoma and in another with pulmonary sequestration (Figs. 3, 4). We hypothesized that tumour uptake of [99mTc]Tc-Galacto-RGD2 depended on the neo-vasculature and integrin αvβ3 expression in the tumour cells, and focal uptake in RGD-targeted imaging would thus be higher in primary tumours with more neo-vasculature and higher integrin αvβ3 expression in the cell membrane. Regarding the different subtypes of lung cancer, LSC had more neovascularization and higher integrin αvβ3 expression, followed by LAC, while SCLC showed less neovascularization and a higher proliferation index. The highest T/N ratio was therefore found in LSC, and was significantly higher than that in LAC and SCLCs. RGD-targeted imaging may thus serve as a useful tool for the phenotyping of lung cancer.[68 Ga]Ga and [18F]F labelled RGD tracers have been used in the preclinical and clinical, [68 Ga]Ga-NODAGA-RGD provide a different spatial distribution than 2-[18F]FDG. It is worth noting that [18F]F-Galacto-RGD not only can be used for the assessment of αvβ3 expression in the tumour neovasculature, but also in humanatherosclerotic carotid plaques, where it correlates with αvβ3 expression [36]. Compared with PET RGD tracer, [99mTc]Tc-Galacto-RGD2 SPECT imaging has a disadvantage in space resolution. However, if taking the expenditure into account, [99mTc]Tc-Galacto-RGD2 has a significant advantage, and the one-kit vial of [99mTc]Tc-Galacto-RGD2 makes the synthesis more convenient, both of them contribute to the clinical transformation and application of [99mTc]Tc-Galacto-RGD2.However, there are some limitations in this study should be taken into concern. First, the quantitation of integrin αvβ3 expression in the immunohistochemistry could be influenced by tumour specimen obtaining and vision field selection. Second, tumour specimens were achieved only in 66 patients, not in all suspected patients, which might influence the data analysis.
Conclusions
This was the first extensive longitudinal study to investigate the expression of integrin αvβ3 in lung cancer. [99mTc]Tc-Galacto-RGD2 imaging showed high sensitivity for the detection of primary lung cancer, but limited specificity. [99mTc]Tc-Galacto-RGD2 uptake in the primary tumour was attributed to integrin αvβ3 expression in the endothelial and tumour cells, and focal uptake occurred in primary lung cancers with more neovascularization and high levels of αvβ3 in the tumour cells. LSC had a higher density of neo-vessels and higher αvβ3 expression, followed by LAC and then SCLC, advanced lung cancer showed higher levels of integrin αvβ3 compared with early stage. These findings suggest that RGD-based imaging might be a useful tool for lung cancer phenotyping and tumour biological behaviour evaluation. Further studies are warranted to validate these findings.
Authors: Eva Hoster; Andreas Rosenwald; Françoise Berger; Heinz-Wolfram Bernd; Sylvia Hartmann; Christoph Loddenkemper; Thomas F E Barth; Nicole Brousse; Stefano Pileri; Grzegorz Rymkiewicz; Roman Kodet; Stephan Stilgenbauer; Roswitha Forstpointner; Catherine Thieblemont; Michael Hallek; Bertrand Coiffier; Ursula Vehling-Kaiser; Réda Bouabdallah; Lothar Kanz; Michael Pfreundschuh; Christian Schmidt; Vincent Ribrag; Wolfgang Hiddemann; Michael Unterhalt; Johanna C Kluin-Nelemans; Olivier Hermine; Martin H Dreyling; Wolfram Klapper Journal: J Clin Oncol Date: 2016-02-29 Impact factor: 44.544
Authors: Shundong Ji; Andrzej Czerwinski; Yang Zhou; Guoqiang Shao; Francisco Valenzuela; Paweł Sowiński; Satendra Chauhan; Michael Pennington; Shuang Liu Journal: Mol Pharm Date: 2013-08-06 Impact factor: 4.939