Abstract
CXCR4 is a chemokine receptor that is overexpressed in various human cancers and is involved in tumor metastasis. The aim of this proof-of-concept study was to evaluate a novel CXCR4-targeted PET probe in patients with solid cancers with reported in vitro evidence of CXCR4 overexpression and to estimate its potential diagnostic value. Methods: Twenty-one patients with histologically proven pancreatic cancer, laryngeal cancer, non–small cell lung cancer, prostate cancer, melanoma, breast cancer, hepatocellular carcinoma, glioblastoma, sarcoma, or cancer of unknown primary underwent PET imaging using the novel CXCR4 nuclear probe 68Ga-pentixafor. The SUVmax of the liver, spleen, and bone marrow was measured to determine physiologic tracer distribution. For evaluation of tracer accumulation in solid cancers, SUVmax and tumor-to-background (T/B) ratios were determined in a total of 43 malignant lesions, including 8 primary tumors, 3 locally recurrent tumors, and 32 metastases. When available, the SUVmax of malignant lesions was compared with the corresponding SUVmax measured in routine 18F-FDG PET. Results: Moderate tracer accumulation was detectable in the liver, bone marrow, and spleen, with a mean SUVmax of 3.1, 3.7, and 5.6, respectively. By visual interpretation criteria, 9 of 11 primary and locally recurrent tumors were detectable, exhibiting a mean SUVmax of 4.7 (range, 2.1–10.9) and a mean T/B ratio of 2.9. Twenty of 32 evaluated metastases were visually detectable (mean SUVmax, 4.5 [range, 3.2–13.8]; mean T/B ratio, 2.8). The highest signal was detected in a patient with non–small cell lung cancer (SUVmax, 10.9; T/B ratio, 8.4) and a patient with cancer of unknown primary (SUVmax, 13.8; T/B ratio, 8.1). Compared with 18F-FDG PET, which was additionally performed in 10 patients, 68Ga-pentixafor PET had a lower SUVmax in all measured malignant lesions. Conclusion: On the basis of these first observations in a small and heterogeneous patient cohort, the in vitro CXCR4 expression profile of solid cancers and metastases described in the previous literature does not seem to sufficiently depict the in vivo distribution revealed by CXCR4-targeted PET. Moreover, the detectability of solid cancers seems to be generally lower for 68Ga-pentixafor than for 18F-FDG PET.
CXCR4 is a 7-transmembrane G-coupled receptor belonging to the chemokine receptor family and is expressed by a variety of cells during development and thereafter (1). With its cognate ligand stromal cell–derived factor 1α (2,3), its main role in the hematopoietic system is to control stem cell retention and the homing of hematopoietic cells to the bone marrow (4). In addition to these physiologic roles, CXCR4 has been found to be overexpressed by various human cancers, including breast, lung, colon, and pancreatic cancer; melanoma; and hematopoietic malignancies such as multiple myeloma (1,5–10). Furthermore, a growing body of evidence shows that the CXCR4 stromal cell–derived factor 1α axis plays a crucial role in the mechanism of cancer and metastasis. High levels of stromal cell–derived factor 1α in organs and tissues such as lymph nodes, lung, liver, bone, and bone marrow are thought to direct the metastasis of CXCR4-expressing tumor cells (11,12). Concordantly, the level of CXCR4 expression was shown to be higher in metastatic sites than in the primary tumors (13), and changes in CXCR4 signaling have been shown to significantly alter metastatic burden in animal models (14). Accordingly, CXCR4 overexpression has been identified as an adverse prognostic factor in some malignancies, including non–small cell lung cancer (NSCLC) and breast cancer (6,15). Because of its critical role in cancer, CXCR4 has been designated a potential therapeutic target and several CXCR4 inhibitors have been developed, with some of them reaching clinical trials (2,16).
These observations highlight and support the possibility that noninvasive imaging of CXCR4 may become a highly interesting new diagnostic or prognostic imaging biomarker through the in vivo quantification of CXCR4 expression levels in tumors and the identification of aggressive subpopulations such as postchemotherapy patients. In addition, CXCR4 imaging may serve as a tool for monitoring novel CXCR4-targeted treatment options, such as pharmacologic or endoradiotherapeutic interventions.
However, despite the fundamental role of CXCR4 in cancer pathogenesis and metastasis, no clinically suitable method for CXCR4 assessment and in vivo quantification has been approved so far. To meet the clinical need for a highly specific and sensitive tool for in vivo CXCR4 assessment and quantification, 68Ga-pentixafor (formerly known as 68Ga-CPCR4.2), a CXCR4-targeted high-affinity nuclear probe, has been developed and evaluated in small cell lung cancer models for PET by Wester’s group (17,18). Proof-of-concept investigations on lymphoma-xenografted animal models and the first investigations on patients with hematologic malignancies confirmed high tracer accumulation in tumors, low uptake in background, and concomitantly fast tracer clearance from nontargeted tissues (18,19). In addition, whole-body dosimetry studies on humans revealed excellent pharmacokinetics and thus low radiation burden to patients (20). Subsequent PET investigations on mice xenografted with human CXCR4-positive multiple myeloma cell lines and patients with advanced multiple myeloma provided high-contrast images with higher detection rates for multiple myeloma manifestations than can be achieved with standard 18F FDG PET/CT scans (21).
On the basis of these highly encouraging data, we studied the use of 68Ga-pentixafor PET in patients with solid cancers and here present our first results. We selected malignancies with previously reported in vitro evidence of CXCR4 overexpression to evaluate whether this imaging modality might be of diagnostic value in CXCR4-expressing cancers. Specifically, we intended to correlate tracer accumulation with reported in vitro expression patterns.
The signal intensities of 68Ga-pentixafor PET in physiologic and tumor tissue were assessed by SUVmax measurements and were compared with those of 18F-FDG PET, when available.
MATERIALS AND METHODS
Patients
Between 2013 and 2014, 21 patients (mean age, 63 y; range, 55–85 y) with histologically proven solid cancers underwent either 68Ga-pentixafor PET/CT or 68Ga-pentixafor PET/MR. Detailed patient characteristics, including cancer type and therapeutic status, are given in Table 1. Ten patients additionally underwent a diagnostic 18F-FDG PET/CT study for staging purposes within 2 wk after 68Ga-pentixafor imaging. No therapy was performed in the interim.
68Ga-Pentixafor PET Results for All Patients
68Ga-pentixafor was synthesized and administered according to the requirements of the German Medicinal Products Act (Arzneimittelgesetz §13 2b) and with the approval of the responsible local regulatory authority (Regierung von Oberbayern). Before the investigation began, all patients gave written informed consent to participate. The data analysis was approved by the responsible local ethics committees.
Synthesis of 68Ga-Pentixafor
68Ga-pentixafor was synthesized using a fully automated, good-manufacturing-practice–compliant procedure on a good-radiopharmaceutical-practice module (Scintomics GmbH) equipped with disposable single-use cassette kits, as previously described (17,19). Before injection, tracer quality was assessed according to the standards of the European Pharmacopoeia (www.edqm.eu). The radiochemical purity of the ready-to-inject formulation was always greater than 99% as confirmed by radio–high-performance liquid chromatography and radio–thin-layer chromatography. The specific activity was in the range of 30–65 GBq/μmol.
PET Imaging Protocol
All 18F-FDG PET scans and 15 of the 21 68Ga-pentixafor scans were obtained on a PET/CT device (Sensation 64 Biograph; Siemens Medical Solutions); the other 6 68Ga-pentixafor scans were obtained on a PET/MRI device (Biograph mMR; Siemens Medical Solutions).
For PET/CT, a low-dose CT scan (26 mAs, 120 kV, 5-mm slice thickness) from the base of the skull to the mid thigh was acquired for attenuation correction, followed by the PET acquisition and the diagnostic CT acquisition (240 mAs, 120 kV, 5-mm slice thickness) in the portal venous phase. The injected activity was 147–275 MBq for 68Ga-pentixafor and 198–359 MBq for 18F-FDG. The PET acquisition began at a mean of 62 min after 68Ga-pentixafor injection and 75 min after 18F-FDG injection. All scans were acquired in 3-dimensional mode at 3 min per bed position. The images were reconstructed using attenuation-weighted ordered-subsets expectation maximization (4 iterations, 8 subsets) followed by a smoothing gaussian filter (5 mm in full width at half maximum).
For PET/MR, the acquisition began at a mean of 72 min after 68Ga-pentixafor injection. A coronal 2-point Dixon 3-dimensional volumetric interpolated T1-weighted sequence was performed to generate attenuation maps (22), followed by a dedicated diagnostic sequence for the type of malignancy being examined. The PET data were acquired simultaneously with the MR data in 3-dimensional mode at 3 min per bed position.
Image Analysis
All 68Ga-pentixafor PET/CT and PET/MR images were interpreted by a board-certified nuclear medicine physician and a board-certified radiologist in a binary fashion. Lesions were defined as visually detectable if they could be seen by both reviewers on the PET images. Semiquantitative analysis with determination of SUVmax involved drawing regions of interest around the primary tumor (in the case of preoperative imaging) and around the metastases (when detectable). If patients had more than 5 metastases, the five with the highest visual uptake were analyzed. The analysis included those lesions that either were histologically verified or were unambiguously malignant because of their morphologic characteristics and dynamic behavior on follow-up. To reduce partial-volume effects, lesions smaller than 10 mm in diameter were excluded from the analysis. For patients who underwent imaging with both 18F-FDG PET/CT and 68Ga-pentixafor PET/CT, analysis of the two sets of images was performed analogously. SUVmax was compared between the 18F-FDG and 68Ga-pentixafor PET images and was tested for a correlation. With reference to Drzezga et al. (22), we did not differ between the SUVmax of PET/MR and PET/CT images, assuming that the semiquantitative values of the two modalities would be comparable.
Statistics
Statistical analysis was performed using MedCalc, version 10.2. We used the Mann–Whitney U test to determine the significance of differences in variables between patient subgroups, and we used the Spearman rank test to check for a correlation in SUVmax between 68Ga-pentixafor PET and 18F-FDG PET.
RESULTS
Patient Characteristics
We examined 21 patients with histologically proven solid malignancies: pancreatic cancer, laryngeal cancer, NSCLC, prostate cancer, melanoma, breast cancer, hepatocellular carcinoma, glioblastoma, sarcoma, and cancer of unknown primary (Table 1). Of the 9 patients who had not received therapy before imaging, a primary tumor was present in eight. Of the 12 patients who had received therapy (chemotherapy, radiotherapy, or surgery), the primary tumor had locally recurred in three. One patient had CUP syndrome for which imaging had been unable to identify the primary tumor. Evidence of metastases was seen in 13 patients, three of whom showed more than 5 metastases each.
Physiologic Tracer Accumulation
A moderate accumulation of 68Ga-pentixafor occurred in the liver and bone marrow, with a mean SUVmax of 3.1 (range, 1.5–4.1) and 3.7 (range, 1.52–6.2), respectively. The spleen demonstrated the highest mean 68Ga-pentixafor positivity (mean SUVmax, 5.6; range, 2.5–10.5) (Fig. 1). There was no significant correlation in SUVmax between treated and nontreated patients for any organ.
Maximum-intensity-projection PET image after tumor excision demonstrates physiologic distribution of 68Ga-pentixafor. Highest physiologic uptake is observed in urinary tract (because of renal excretion) and in spleen. Tracer accumulation at right elbow reflects injection site.
68Ga-Pentixafor Accumulation in Solid Tumors
Forty-three malignant lesions were evaluated in 21 patients (8 primary tumors, 3 locally recurrent tumors, and 32 metastases; Table 1). By visual interpretative criteria, 9 of 11 primary and locally recurrent tumors were detectable, exhibiting a mean SUVmax of 4.7 (range, 2.1–10.9) and a mean tumor-to-background (T/B) ratio of 2.9 (range, 0.2–8.4). The highest SUVmax, 10.9, was measured in a patient with NSCLC. Twenty of 32 evaluated metastases were visually detectable (mean SUVmax, 4.5 [range, 3.2–13.8]; mean T/B ratio, 2.8 [range 1.3–8.1]). The highest SUVmax, 13.8, was in the cervical metastases of the patient with cancer of unknown primary. The Spearman test revealed a low correlation between SUVmax and number of lesions per patient (r = 0.3).
Comparison of 68Ga-Pentixafor PET with 18F-FDG PET
18F-FDG PET had been performed within 2 wk after 68Ga-pentixafor PET for 10 patients, with a total of 27 lesions evaluated (2 primary tumors and 25 metastases) (Table 2). All 27 lesions were visually detectable with 18F-FDG, but only 19 with 68Ga-pentixafor. Semiquantitative analysis revealed that SUVmax and T/B ratio in all measured lesions were significantly higher for 18F-FDG than for 68Ga-pentixafor (Figs. 2 and 3). Accordingly, the mean SUVmax of all lesions was significantly higher on 18F-FDG images than on 68Ga-pentixafor images (SUVmax, 13.8 vs. 5.0; P > 0.001). Details of the per-patient analysis are presented in Table 2. The Spearman test revealed a low correlation in SUVmax between 18F-FDG PET and 68Ga-pentixafor PET (r = 0.3; P = 0.99).
Comparison of 68Ga-Pentixafor and 18F-FDG PET Results in the Patients Who Underwent Both
50-y-old patient with histologically proven invasive ductal carcinoma of right breast. Tumor demonstrates intensive uptake on transversal (A), coronal (B), and fusion (C) 18F-FDG PET/CT images. In comparison, uptake is significantly lower on transversal (D), coronal (E), and fusion (F) 68Ga-pentixafor PET/CT images.
50-y-old patient with invasive ductal carcinoma of right breast and diagnosis of bone metastasis (arrows). Metastasis demonstrates intensive focal uptake on transversal (A), coronal (B), and fusion (C) 18F-FDG PET/CT images. In comparison, uptake is not visually detectable on transversal (D), coronal (E), and fusion (F) 68Ga-pentixafor PET/CT images.
DISCUSSION
CXCR4 is a key chemokine receptor involved in many processes during tumorigenesis and critical for cancer dissemination (1,5–10). In this study, we estimated the diagnostic potential of CXCR4-targeted PET imaging in solid cancers for which in vitro evidence of CXCR4 overexpression has been reported (Table 1). For example, significant CXCR4 overexpression has been demonstrated in pancreatic cancer cells compared with normal tissue (23), and high-intensity staining for CXCR4 has been observed in chondrosarcomas through immunohistochemistry, with the most pronounced expression profile being in high-grade sarcomas. Similarly, the aggressiveness of a glioblastoma seems to correlate with its CXCR4 expression pattern (25), and a high accumulation of CXCR4 has been reported to be especially prevalent in breast carcinomas of high metastatic potential (5,6). In melanoma and NSCLC, the detection of high CXCR4 expression has even led to the preclinical development of antagonists, with promising results (26,27).
Interestingly, despite the reported in vitro evidence of CXCR4 overexpression in these solid cancers, 68Ga-pentixafor positivity varied significantly between different tumor entities, with an SUVmax of 10 being exceeded in only 2 patients (one who had NSCLC with a T/B ratio of 8.4 and one who had cervical metastases from cancer of unknown primary with a T/B ratio of 8.1). All other solid malignancies, including pancreatic carcinoma, laryngeal carcinoma, hepatocellular carcinoma, melanoma, breast cancer, glioblastoma, and sarcoma, demonstrated low to modeate 68Ga-pentixafor positivity. An accumulation of 68Ga-pentixafor was visually detectable in only 29 of 43 cancer lesions—both primary tumors and metastases.
Different reasons might account for the discordance between the high CXCR4 expression profile described in the literature and our current findings. The level of CXCR4 expression assessed by either transcript or whole-cell protein level analysis is not necessarily representative of the CXCR4 expression level on the cell surface (10,28). Shim et al., for example, demonstrated that CXCR4 expression in lymph node metastases of breast cancer patients occurs predominantly in the cytoplasm (28). Because 68Ga-pentixafor PET binds to membrane-associated chemokine receptors, a significant discordance might occur between CXCR4 expression profiles determined by transcript or whole-cell protein level analysis and by in vivo quantification of CXCR4 using PET probes. Further relevant factors in this setting might involve the overexpression of CXCR4 on cancer stem cells, which are believed to represent a drug-resistant cell population that can survive even chemotherapy (29–32). Interestingly, the number of cancer stem cells seems to differ with the molecular subtype of the tumor (33) and is believed to correlate with tumor aggressiveness and metastatic potential.
We compared the diagnostic potential of 68Ga-pentixafor PET with that of 18F-FDG PET by evaluating the 10 of 21 patients with solid malignancies who underwent imaging with both modalities. All 27 of the evaluated malignant lesions within this subpopulation were visually detectable on 18F-FDG PET, compared with only 19 on 68Ga-pentixafor PET. Moreover, SUVmax in all lesions was higher on 18F-FDG PET than on 68Ga-pentixafor PET (SUVmax, 13.8 vs. 5.0; P < 0.001). Our results indicate that the detectability of solid cancers may generally be lower on 68Ga-pentixafor PET than on 18F-FDG PET. These preliminary findings indicate that, for use as a standard oncologic tool, CXCR4-targeted PET does not seem as suitable as 18F-FDG PET. However, the potential of CXCR4-targeted PET might be found in certain specialized applications yet to be elucidated. Several other receptor-binding PET imaging probes already in routine clinical use are tailored to specific tumor types and characteristics. Apart from the often-used somatostatin receptor ligands DOTATATE, DOTATOC, DOTANOC, and HA-DOTATATE for imaging neuroendocrine tumors with low proliferation rates, ligands that bind with high affinity to prostate-specific membrane antigen have been proven to significantly enhance the specificity of prostate cancer imaging (34). Similarly, imaging of CXCR4 might be valuable in only certain subtypes of solid tumors. For example, the benefit of CXCR4 PET imaging in a subset of multiple myeloma patients was recently demonstrated by Philipp-Abbrederis et al., who found specificity and contrast to be superior to 18F-FDG (21).
Future efforts should focus on characterization of additional cancer types in which CXCR4 imaging might demonstrate a diagnostic benefit. Besides having the potential to allow high-specificity imaging in certain cancer types, the CXCR4 expression profile would be highly interesting when considering 68Ga-pentixafor as a selection marker for CXCR4-directed treatment. 177Lu- or 90Y-coupled pentixafor analogs could become attractive radiopharmaceuticals for a theranostic approach, with 68Ga-pentixafor PET being used as a marker for patient selection.
The only moderate 68Ga-pentixafor positivity of the liver and the bone marrow demonstrated in this study might indicate a favorable side effect profile. The significance of the high variance in splenic tracer uptake, however, has to be further elucidated.
One major limitation of the current study design is the heterogeneity of the patient cohort and the low number of patients per tumor entity. We included patients with various types of malignancies and various stages of disease. A further limitation is the lack of immunohistochemical cross validation. However, we intended to provide a cross section of different cancer entities that have been described in the literature as having considerable CXCR4 expression. The general imaging characteristics of 68Ga-pentixafor, such as high specificity of binding, low nonspecific binding, and low background uptake, have already been demonstrated in first patient studies (21). Consequently, our study can serve only as a first impression on the suitability of 68Ga-pentixafor PET imaging for patients with the solid tumors listed in Table 1. Such an approach is plausible, considering that experience with this new target and tracer is extremely limited and the areas of application still need to be found. For example, on the basis of the observed findings, examination of NSCLC patients with CXCR4-targeted tracers might be worth investigating. Moreover, metastases from cancer of unknown primary exhibited the highest tracer uptake of all examined lesions, suggesting a potential diagnostic benefit in the detection of the primary cancer. For certain tumor entities, dedicated prospective clinical studies that include correlations with biopsied tumor samples have been initiated.
CONCLUSION
Our data demonstrate the feasibility of 68Ga-pentixafor for PET imaging of solid malignancies. The cancers seemed to exhibit a heterogeneous 68Ga-pentixafor accumulation that was lower than that of 18F-FDG. Moreover, the previously described in vitro CXCR4 expression profiles of solid cancers and metastases do not seem to sufficiently depict the in vivo distribution revealed by CXCR4-targeted PET, making further research necessary. However, once the areas of pentixafor imaging are more clearly defined, PET imaging of CXCR4 may prove a valuable modality either as a stand-alone diagnostic tool or in combination with 18F-FDG PET; that is, when considering 68Ga-pentixafor for monitoring CXCR4-directed pharmacologic or endoradiotherapeutic treatment.
DISCLOSURE
The costs of publication of this article were defrayed in part by the payment of page charges. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734. This work received support from the ERC grant MUMI, from the Deutsche Forschungsgemeinschaft (DFG) under grant agreement SFB 824, and from the German Jose Carreras Leukemia Foundation (grant R11/18). Hans-Jürgen Wester is the CEO of Scintomics, the distributor of pentixafor. No other potential conflict of interest relevant to this article was reported.
Footnotes
Published online Jan. 14, 2016.
- © 2016 by the Society of Nuclear Medicine and Molecular Imaging, Inc.
REFERENCES
- Received for publication May 22, 2015.
- Accepted for publication December 1, 2015.
Jump to section
Related Articles
Cited By...
- Imaging of C-X-C Motif Chemokine Receptor 4 Expression in 690 Patients with Solid or Hematologic Neoplasms Using 68Ga-Pentixafor PET
- 68Ga-Pentixafor PET/CT Demonstrating In Vivo CXCR4 Receptor Overexpression in Rare Lung Malignancies: Correlation with Histologic and Histochemical Findings
- CXCR4-Targeted PET Imaging of Central Nervous System B-Cell Lymphoma
- Imaging Inflammation in Atherosclerosis with CXCR4-Directed 68Ga-Pentixafor PET/CT: Correlation with 18F-FDG PET/CT
- 64Cu PET Imaging of the CXCR4 Chemokine Receptor Using a Cross-Bridged Cyclam Bis-Tetraazamacrocyclic Antagonist
- Modulators of CXCR4 and CXCR7/ACKR3 Function
- Feasibility of CXCR4-Directed Radioligand Therapy in Advanced Diffuse Large B-Cell Lymphoma
- CXCR4 Is a Potential Target for Diagnostic PET/CT Imaging in Barrett's Dysplasia and Esophageal Adenocarcinoma
- Glu-Ureido-Based Inhibitors of Prostate-Specific Membrane Antigen: Lessons Learned During the Development of a Novel Class of Low-Molecular-Weight Theranostic Radiotracers
- CXCR4 Ligands: The Next Big Hit?