In vivo molecular imaging of chemokine receptor CXCR4 expression in patients with advanced multiple myeloma

Abstract

CXCR4 is a G-protein-coupled receptor that mediates recruitment of blood cells toward its ligand SDF-1. In cancer, high CXCR4 expression is frequently associated with tumor dissemination and poor prognosis. We evaluated the novel CXCR4 probe [(68)Ga]Pentixafor for in vivo mapping of CXCR4 expression density in mice xenografted with human CXCR4-positive MM cell lines and patients with advanced MM by means of positron emission tomography (PET). [(68)Ga]Pentixafor PET provided images with excellent specificity and contrast. In 10 of 14 patients with advanced MM [(68)Ga]Pentixafor PET/CT scans revealed MM manifestations, whereas only nine of 14 standard [(18)F]fluorodeoxyglucose PET/CT scans were rated visually positive. Assessment of blood counts and standard CD34(+) flow cytometry did not reveal significant blood count changes associated with tracer application. Based on these highly encouraging data on clinical PET imaging of CXCR4 expression in a cohort of MM patients, we conclude that [(68)Ga]Pentixafor PET opens a broad field for clinical investigations on CXCR4 expression and for CXCR4-directed therapeutic approaches in MM and other diseases.

Keywords: CXCR4; chemokine receptor; in vivo imaging; multiple myeloma; positron emission tomography.

Figure 1. CXCR4 expression in MM patient bone marrow

1. Flow cytometric evaluation of CXCR4 surface expression using an anti-CXCR4-PE antibody. Left: positive patient; right: negative patient; representative data are shown. The gating strategy is depicted in Supplementary Fig S1.

2. Representative histograms revealing the magnitude of CXCR4 expression in myeloma cells as compared to the indicated bone marrow cell subtype.

3. Median fluorescence intensity of surface CXCR4 expression relative to isotype control (n = 7–10 patients). Horizontal bars indicate mean of all individual patient values; asterisks indicate statistical significance (plasma cells vs B cells: P = 0.0082; plasma cells vs T cells: P = 0.0091; plasma cells vs CD34⁺ cells: P = 0.011; plasma cells vs monocytes: P = 0.0423; Student's t-test to compare mean plasma cell relative expression with mean relative expression of each further indicated cell subtype). MM patients judged positive for CXCR4 expression were selected for this analysis.

4. Representative MM bone marrow staining of a patient positive for MM CXCR4 expression: hematoxylin and eosin (H&E) staining. Immunohistochemistry for CD138, CXCR4, light chain kappa, and light chain lambda.

Figure 2. [⁶⁸Ga]Pentixafor PET imaging of MM xenografts

1. Real-time PCR analysis of cxcr4 transcript expression levels in HeLa (negative control) and in MM cell lines MM.1S and OPM-2. Shown is the mean relative expression ± SEM (n = 3 independent experiments). Values are normalized to the expression of ubiquitin (Ub). The asterisk indicates statistically significant differences (HeLa vs MM.1S: P < 0.001; HeLa vs OPM-2: P < 0.0001; Student's t-test).

2. CXCR4 protein expression assessed by immunoblotting (one representative blot out of 3 is shown).

3. Binding of [⁶⁸Ga]Pentixafor to MM.1S and OPM-2 cells after the indicated incubation periods (n = 3 per cell line and time point). Shown is the mean ± SEM. The difference between the OPM-2 groups is statistically significant; *P < 0.0017 (one-way ANOVA).

4. Mean tumor-to-background ratio (TBR) for [¹⁸F]FDG (left) and for [⁶⁸Ga]Pentixafor (right) in MM.1S and OPM-2 xenograft-bearing NOD SCID mice. Shown is the mean ± SEM, n = 8 tumors (4 mice); *P = 0.0111 for [¹⁸F]FDG and *P = 0.0113 for [⁶⁸Ga]Pentixafor (Student's t-test). One-way ANOVA revealed significant differences between the groups; P < 0.0001 (not graphically shown).

5. Representative [⁶⁸Ga]Pentixafor PET images of three mice bearing MM.1S (right shoulder) and OPM-2 (left shoulder) tumors.

6. Flow cytometric quantification of CXCR4 cell surface expression on resected MM.1S and OPM-2 tumors. Data are the mean ± SEM, n = 4. *P = 0.0286; Mann–Whitney U-test.

7. Correlation of [⁶⁸Ga]Pentixafor PET mean TBR and CXCR4 cell surface expression assessed by flow cytometry. n = 8 tumors were analyzed.

8. Mice (n = 4) bearing OPM-2 and MM.1S xenografts were coinjected with AMD3100 (right image, one representative mouse) or not pretreated (left image) before undergoing [⁶⁸Ga]Pentixafor PET. The white arrows point to the bladder. Quantification is shown in Supplementary Fig S3A.

Source data are available online for this figure.

Figure 3. [⁶⁸Ga]Pentixafor PET/CT and [¹⁸F]FDG PET/CT

1. A–D Maximum intensity projections (MIP) of [⁶⁸Ga]Pentixafor (A) and [¹⁸F]FDG PET/CT (B) of a 68-year-old male with histologically proven multiple myeloma indicating the better lesion-to-background contrast for [⁶⁸Ga]Pentixafor in the corresponding myeloma manifestations. Trans-axial views of the upper thorax (C) and the pelvis (D) underline the higher uptake values of the bone manifestations (yellow arrows) of [⁶⁸Ga]Pentixafor compared to [¹⁸F]FDG.

Figure 4. Visual comparison of [¹⁸F]FDG- and [⁶⁸Ga]Pentixafor PET scans

1. Number of patients with visual positivity for the indicated PET tracer (total: n = 14).

2. Number of patients (total n = 14) for whom imaging with [¹⁸F]FDG PET (FDG, n = 2) or [⁶⁸Ga]Pentixafor PET (Pentixafor, n = 7) was superior, with comparable positivity (comparable, n = 3), and with dual imaging providing complementary visual information (complementary, n = 2).

Figure 5. [⁶⁸Ga]Pentixafor PET/MR images

1. A–C Coronal views of [⁶⁸Ga]Pentixafor (A), T2 STIR weighted MRI (B) and CT bone window (C) of a male patient with histologically proven multiple myeloma. The increased [⁶⁸Ga]Pentixafor uptake correlates with the hyperintense T2 STIR signal; however, the myeloma manifestations are underestimated in the corresponding bone window CT.