Clinical study of 99mTc-3P-RGD2 peptide imaging in osteolytic bone metastasis

Objective To investigate the value of integrin αvβ3 targeted imaging with 99mTc-HYNIC-PEG4-E[PEG4-c(RGDfk)]2 (99mTc-3P-RGD2) as a radiotracer in dectecting osteolytic bone metastases. Methods This is a retrospective study involving a cohort of 69 consecutive patients including 59 with lung cancer and 10 with other cancers. Patients were required to receive whole body scan (WBS) and regional SPECT/CT imaging with 99mTc-3P-RGD2 (RGD imaging) and 99mTc-MDP (MDP imaging) as a radiotracer successively within days. Final diagnosis was based on comprehensive assessment of all available data including case history, CT, MRI, SPECT/CT, PET/CT, histopathology and 6-12 months follow-up. Visual observation and semiquantitative analysis (T/N: tracer uptake ratio of osteolytic metastases to normal bone) of 99mTc-3P-RGD2 or 99mTc-MDP imaging were performed and their detective values for osteolytic metastases were compared. Results A total of 131 osteolytic metastatic lesions were retrospectively studied. Osteolytic metastases mainly presented as “hot region”, occasionally as “cool or normal region” on RGD imaging. The detection sensitivity of RGD WBS for osteolytic metastases was significantly higher than that of 99mTc-MDP WBS (80.9% vs. 46.6%, p<0.01). The sensitivity increased to 96.2% (126/131) when combining with SPECT/CT. 99mTc-3P-RGD2 imaging also promoted the detection of unknown primary tumor, lymph node metastases and offered information for clinical staging. T/N of 99mTc-3P-RGD2 in lung adenocarcinoma osteolytic metastases showed no statistical difference compared with that in squamous-cell carcinoma (6.84±3.46 vs. 7.33±3.22, t = 0.39, p = 0.71). Whereas, it was higher in osteolytic metastases from lung cancer than that from thyroid cancer (7.05±3.01 vs. 4.11±2.67, p = 0.03). Conclusion 99mTc-3P-RGD2 peptide imaging showed great potential for detection of osteolytic bone metastasis due to high expression level of integrin αvβ3 on osteoclast and most tumor cells.


INTRODUCTION
Skeleton is one of the most common sites for metastasis. A majority of patients may present osteolytic and osteoblastic metastasis, which will result in marked disturbances of bone remodeling that can be lytic and/or blastic. Osteolytic metastasis, a common complication in breast cancer, lung cancer, prostate cancer, or multiple Clinical Research Paper myeloma [1,2], brings about significant morbidity of intractable pain, spinal cord compression, pathologic fracture, functional impairment and hypercalcemia [3,4]. In addition, bone micro-metastasis can serve as an independent predictor of poor outcome in patients with tumor even among lymph node-negative patients with primary tumors of less than 2 cm [5]. Early diagnosis of osteolytic metastasis is definitely important for its correct treatment and better prognosis.
Anatomical imaging (e.g. CT and MRI) has been commonly used for the diagnosis of osteolytic lesions. However, the diagnostic sensitivity of CT is relatively low in cases of a mineral content loss of less than 50% at the lesions sites [6]. Meanwhile, MRI could not readily display bone degradation [7]. 99m Tc-diphosphonates ( 99m Tc-MDP)-based planar bone scan is considered as the standard technique for the detection of skeletal metastasis as it shows high sensitivity, especially for bone lesions with osteogenesis. As is known to all, the accumulation of 99m Tc-MDP in the bone is highly relied on the osteoblast activity [8]. Absence of 99m Tc-MDP or high false negative was reported at early period in osteolytic lesions rich in osteoclast, tumor cells and different degree of osteolysis. Therefore, the detection value of 99m Tc-MDP imaging in osteolytic lesions was lower than that in the osteoblastic lesions [9]. On this basis, it is urgent to develop appropriate imaging technique(s) for the detection of osteoclasts and adjacent tumor cells in patients with osteolytic metastases in an early and accurate manner.
Malignant interaction between tumor cells and osteoclasts in bone microenvironment plays a pivotal role in the pathogenesis of metastatic bone disease such as osteolytic and/or osteoblastic bone metastases. Tumor cells contributed to the recruitment and activation of osteoclasts, which triggered regional osteolysis and tumor cell proliferation [10]. Tumor cells and osteoclasts are the main component of osteolytic metastatic lesions. Integrin α v β 3 , highly expressed in several tumor cells and activated endothelial cells in newly-generated vessels, has been considered as a target for tumor imaging with radiolabeled arginine-glycine-aspartic acid (RGD) peptides and analogues. Furthermore, osteoclast expressing high level of integrin α v β 3 had attracted great attention for positive imaging of osteolytic bone metastasis [11]. ( 99m Tc-3P-RGD 2 ) is a cyclic RGD dimmer peptide with high specificity and affinity to integrin α v β 3 . To date, 99m Tc-3P4-RGD 2 scintigraphy has been commonly used for differential diagnosis of solitary pulmonary nodule, lymph node metastasis and treatment response monitoring [12][13][14][15]. This retrospective study was designed to investigate the detective and diagnostic value of 99m Tc-3P4-RGD 2 imaging for osteolytic bone metastasis.

Patients
This study was approved by the Institute Review Boards of both Nanjing Medical University and Nanjing First Hospital. Written informed consent was obtained from each patient. All reported investigations were conducted in accordance with the Declaration of Helsinki and with our national regulations. Eighty-eight patients were diagnosed with malignant tumors such as primary lung cancer, lung metastases from thyroid cancer, malignant chromaffin-cell tumor, gastric cancer, breast cancer based on the assessment of case history, CT, MRI, SPECT/CT, PET/CT histopathology and 6-12 month follow-up data. Sixty-nine (48.6%) patients presenting pulmonary nodule were suspected with lung cancer and concurrent osteolytic bone metastases. The exclusion criteria were as follows: (i) received treatment before imaging; (ii) pregnant and breastfeeding patients; (iii) those with a history of bone trauma, fracture, and bone inflammation such as tuberculosis within one year; or (iv) those could not accomplish the required examinations because of severe pain or claustrophobia.

Imaging protocol
We evaluated all of the patients using 99m Tc-3P-RGD 2 imaging and 99m Tc-MDP imaging [including both whole body scan (WBS) and SPECT-CT with consistent imaging field] within one week. The imaging was performed on a SPECT-CT equipment (Symbia T6, Simense, Germany) according to manufacturer's instructions. All patients had urinated completely prior to imaging. Imaging was performed within one week. Anterior and posterior WBS was performed 1h after intravenous injection of 99m Tc-3P-RGD 2 (750±37.5MBq, www.impactjournals.com/oncotarget in 1.0 ml saline) or 3h after intravenous injection of 99m Tc-MDP (750±75MBq, in 1.0 ml saline). WBS imaging parameters included low-energy high-resolution collimators, bed movement speed of 10.0 cm/min, energy window of 20% width and centered on 140keV. SPECT data set was obtained (ZOOM of 1.3, 256 by 256 matrix size, 30s/Frame for 32 frames and 180° each head of the dual head-camera) immediately after WBS using a lowenergy high-resolution collimator at 140 keV with a window width of 20%. Then the data were reconstructed with ordered subset expectation maximization (OSEM). CT scan was performed in the same anatomic locations as SPECT with a tube voltage of 130 kV and current intensity at 120 mA/slice with a slice thickness of 3 mm.

Image analysis
Image analysis was performed by 2 nuclear medicine physicians and 2 radiologists blinded to the case history, examination results, and pathologic diagnosis or follow up data. Visual observation and semiquantitative analysis of 99m Tc-3P-RGD 2 or 99m Tc-MDP imaging were performed based on the accumulation of radiotracer at the lesion sites comparing to the contralateral or surrounding normal bone tissues (T/N ratio). The 4-point grade system was also adopted to describe the uptake degree of radiotracers in osteolytic lesions [18]: grade 0, tracer uptake similar to surrounding normal bone structure; grade 1, uptake less ( 99m Tc-MDP) or slightly higher ( 99m Tc-3P-RGD 2 ) than surrounding normal bone structure; grade 2, uptake significantly less ( 99m Tc-MDP) or higher ( 99m Tc-3P-RGD 2 ) than surrounding normal bone structure; grade 3, abnormal aggregation of 99m Tc-3P-RGD 2 or almost absence of 99m Tc-MDP with or without regional accumulation surrounding the lesion or regional aggregation of 99m Tc-MDP in confirmed osteolytic lesions. Osteolytic bone metastasis was ascertained according to any of the following criteria: tracer uptake score greater or equal to grade 2 on either 99m Tc-3P-RGD 2 or 99m Tc-MDP imaging; tracer uptake score equals to grade 1 on both 99m Tc-3P-RGD 2 and 99m Tc-MDP imaging especially with bone pain symptom, or increased uptake on 99m Tc-MDP imaging but confirmed to be osteolytic on other imaging modalities such as CT (including CT on SPECT-CT imaging), 18 F-FDG PET-CT imaging, and MRI.

Verification of osteolytic bone metastases
The final diagnosis of osteolystic bone metastases was based on histopathology, imaging findings (i.e. CT, SPECT/CT, PET/CT, MRI) and clinical follow-up data using specific standards as follows: (a) histopathologically proven; and (b) Besides primary malignancy diagnosis, CT or magnetic resonance imaging (MRI) results indicated obvious bone destruction without osteogenic imaging performance. (c) Increased range and/or lesions of bone destruction during follow-up [19][20][21]. During the clinical follow-up, multiple imagings were required for the patients, and all patients were followed up at least 6 months. Lesions exhibiting both osteolytic and osteosclerotic changes were verified by either type, depending on the predominant changes [22].

Statistical analysis
Data were expressed as mean ± standard deviation. The number of osteolytic bone metastases detected using 99m Tc-3P-RGD 2 and 99m Tc-MDP was compared. Sensitivity of 99m Tc-3P-RGD 2 imaging and 99m Tc-MDP imaging modality were compared using McNemar test. The lesion detection consistency of the two modalities were compared using the Kappa test. P < 0.05 was considered to be statistically significant.

Patient characteristics and distribution of osteolytic lesions
Sixty-nine cases with malignant tumor were confirmed with 131 osteolytic metastases according to the comprehensive data based on case history, CT, MRI, SPECT/CT, PET/CT examinations, histopathology and 6-12 monthes follow-up. Patients' characteristics and the number of osteolytic, lymph node metastases were listed in Table 1.
Based on visual analysis, osteolytic bone metastases were mainly manifested as significant accumulation (hot area, Fig. 1A and 1B), occasionally slight higher uptake (Fig. 1C) or absence of 99m Tc-3P-RGD 2 on WBS and SPECT-CT imaging with or without bone destruction on CT (cold area, Fig. 1D). For 99m Tc-MDP, osteolytic bone metastases were manifested as "cold area" and occasionally as increased uptake (Fig. 2 and 3). Single osteolytic metastasis (with a long diameter of 4.1 cm) was manifested as cold region with slight elevation of 99m Tc-MDP uptake in the peripheral part, while negative findings were observed on 99m Tc-3P-RGD 2 WBS for the single osteolytic metastasis.

Detection value of 99m Tc-3P-RGD2 and 99m Tc-MDP imaging for osteolytic bone metastases
Semiquantification of tracer uptake in osteolytic bone mestastasis was expressed as T/N. A large variance (0.73-13.5) was noticed in the 99m Tc-3P-RGD2 in tracer uptake in osteolytic bone metastases. T/N on 99m Tc-3P-RGD 2 WBS was significantly lower than that on SPECT-CT (4.93±2.20 vs. 6.80±3.19, t = 15.1, p < 0.01). The positive rate of 99m Tc-3P-RGD 2 WBS showed significant increase compared to 99m Tc-MDP WBS based on T/N analysis (80.9% vs. 46.6%, p < 0.05, Supplemental Table  1). When combining with SPECT-CT, the detective sensitivity of 99m Tc-3P-RGD 2 imaging increased to 96.2% (126/131). The osteolytic metastases demonstrated that lower or similar 99m Tc-3P-RGD 2 uptake was more commonly seen in large size lesions with existence of predominant bone destruction, internal necrosis or/and surrounding cyclic hyperostosis/osteosclerosis. T/N of 99m Tc-3P-RGD 2 in osteolytic metastatic lesions in lung adenocarcinoma patients showed no statitical difference from that in squamous-cell carcinoma (6.84±3.46 vs. 7.33±3.22, t = 0.39, p = 0.71). T/N of 99m Tc-3P-RGD 2 in osteolytic metastases from primary lung cancer was significantly higher than that from thyroid cancer    Tc-MDP  99m Tc-3P-RGD 2  axial skeleton  110  60  92  <0.001  appendicular skeleton  21  17  19  >0.05  total  131  77  111  <0.001 collagen in bone matrix. It is manifested as a "hot zone" in osteoblastic bone metastases even before changes of anatomical structure. 99m Tc-MDP WBS imaging is superior for the detection of osteoblastic metastases, however, its efficiency in the detection of osteolytic lesions is hampered due to overlaping from normal bone tissues. Osteolytic bone metastasis, characterized by the activation of osteoclasts and the resulting bone resorption, is visualized as a "cold area" (negative imaging) due to absence of 99m Tc-MDP uptake. However, osteolytic metastatic lesions, consisting of a great number of tumor cells and osteoclasts expressing high level of integrin αvβ3 and αvβ5 [23], are visualized as a "hot zone" (positive imaging) as it can bind to 99m Tc-3P-RGD 2 . This was the main mechanism of 99m Tc-3P-RGD 2 imaging for the detection of osteolytic bone metastases with high T/N ratio. Our study showed that 99m Tc-3P-RGD 2 WBS was more effective than 99m Tc-MDP WBS in the detection of osteolytic bone metastases, which may be related to the following aspects: Firstly, "hot area" was more easy to be visualized than "code area" on WBS. Secondly, distribution of 99m Tc-3P-RGD 2 in normal bone structures was extremely low, which made the osteolytic bone metastases much easier to be detected with low background. Thirdly, overlaping of normal bone structures contributed to low grade for the lesions in vetebrea, sternum, sacrococcyx and pelvis, which finally induced misdiagnosis by 99m Tc-MDP WBS [24].
Clinically, SPECT-CT was performed in presence of abnormality or highly suspicious on WBS images. For the highly suspected patients, it is difficult for a patient to accept the whole body SPECT-CT each time during his first imaging or during regular follow-up. The detection of suspicious osteolytic bone metastases on WBS is of great importance as it serves as a gate for selective SPECT-CT imaging that will facilitate lesion detection, boundary delineation for semiquantitive analysis and accurate diagnosis [25,26]. The detection value of osteolytic bone metastases increased by about 15% and several patients with score grade increased on both imaging modalities after combining with SPECT-CT. SPECT-CT is vital for 99m Tc-3P-RGD 2 imaging due to absence of normal skeleton appearance for lesion localization. Furthermore, moderate-  Tc-MDP  99m Tc-3P-RGD 2  99m Tc-MDP  99m Tc-3P-RGD2  0  54  20  27  10  1  15  29  27  32  2  28  54  39  61  3 34 28 38 28   to-intense 99m Tc-3P-RGD 2 accumulation in visceral organs might make the mild-uptake lesions undetectable in lower thoracic, upper lumber vertebra and sacrococcyx [27]. The diagnostic sensitivity of 99m Tc-3P-RGD 2 imaging for osteolytic bone metastases was significantly higher than that of 99m Tc-MDP imaging in our study. Also, its sensitivity was slightly lower than 18 F-Alfatide II PET-CT imaging [18]. This may be related to the higher spatial resolution of PET and the large database. Besides detection of osteolytic bone metastases, 99m Tc-3P-RGD 2 imaging was also useful for the differential diagnosis of primary tumor and even its staging [14,28]. In our study, primary lung cancer, thyroid cancer, malignant chromaffin-cell tumor and gastric cancer were highly positive on 99m Tc-3P-RGD 2 imaging [29]. Thus, 99m Tc-3P-RGD 2 imaging was superior to 99m Tc-MDP in the detection of osteolytic bone metastases, unknown primary tumor, distant metastases and clinical staging. T/N of 99m Tc-3P-RGD 2 in osteolytic metastasis may be partially influenced by integrin αvβ3 expression level on primary tumor cells [30]. In future, further studies involving more cases of different malignant tumors are needed.
In metastatic bone tissues, 99m Tc-MDP imaging reflects bone metabolism and osteogenesis, while 99m Tc-3P-RGD 2 imaging reflects the existence of integrin αvβ3overexpressing tumor cells, osteoclast, angiogenesis [23,31]. Osteolytic bone metastases usually demonstrated as increased accumulation of 99m Tc-3P-RGD 2 imaging and low uptake of 99m Tc-MDP. However, "hot area" or "cold area" can be visualized on both imaging modalities at the same lesion in this study. The process of osteolytic bone metastases formation is triggered by the interaction among tumor cells, bone marrow environment and bone cells (vicious cycle) [32]. The tissue component changes during the "vicious cycle" until bone destruction and structure absence. "Hot area" on both imaging modalities is possibly due to increased activation of osteolysis and osteogenesis at early stage, while "cold area" is possibly associated with bone structure resorption, destruction, internal necrosis at late stage of osteolytic bone metastasis. The CT information on SPECT-CT is important to distinguish malignant osteolytic bone destruction from benign ones when they demonstrate as "cold area" on both 99m Tc-3P-RGD 2 and 99m Tc-MDP imaging. Thus, these two imaging modalities are potential to play complementary roles in reflecting pathophysiological status of osteolytic lesions. In future, further studies are needed to develop targeted therapies for the management of osteolytic metastasis using 177Lu and 90Y labeled RGD peptide.

CONCLUSIONS
Using integrin αvβ3 highly expressed in the main component of osteolytic bone metastases (e.g. tumor cells and osteoclasts) as a target, 99m Tc-3P-RGD 2 imaging demonstrated mainly as "hot region" on whole body imaging with low background. 99m Tc-3P-RGD 2 imaging showed significantly higher detective rate than 99m Tc-MDP imaging based on visual and semiquantitive analysis. Such technique is potentially applicable for the detection of some unknown primary tumors and distant metastases.