Preliminary evidence of imaging of chemokine receptor-4-targeted PET/CT with [68Ga]pentixafor in non-Hodgkin lymphoma: comparison to [18F]FDG

Background In order to study the CXCR4 expression with [68Ga]pentixafor PET in different types of non-Hodgkin lymphoma, we performed a retrospective study to describe the [68Ga]pentixafor PET/CT imaging in a spectrum of lymphomas and to compare it with [18F]FDG PET/CT. Results Twenty-seven patients with newly diagnosed non-Hodgkin lymphoma were recruited retrospectively. [68Ga]pentixafor PET showed increased radioactivity in lymphoplasmacytic lymphoma (n = 8), marginal zone lymphoma (n = 4), diffuse large B cell lymphoma (n = 3), follicular lymphoma (n = 2), mantle cell lymphoma (n = 1), unclassified indolent B cell lymphoma (n = 3), and enteropathy associated T cell lymphoma (n = 3). However, peripheral T cell lymphoma, not otherwise specified (n = 1), and NK/T cell lymphoma (n = 2) were not avid for [68Ga]pentixafor. In comparison to [18F]FDG PET, [68Ga]pentixafor PET demonstrated more extensive disease and higher radioactivity in lymphoplasmacytic lymphoma and marginal zone lymphoma. Conclusion CXCR4 expression varies in different types of non-Hodgkin lymphoma. Overexpression of CXCR4 was detected with [68Ga]pentixafor PET/CT in lymphoplasmacytic lymphoma, marginal zone lymphoma, diffuse large B cell lymphoma, follicular lymphoma, mantle cell lymphoma, unclassified indolent B cell lymphoma, and enteropathy associated T cell lymphoma. The uptake of [68Ga]pentixafor was higher than [18F]FDG in lymphoplasmacytic lymphoma and marginal zone lymphoma.


Background
C-X-C motif chemokine receptor 4 (CXCR4) is a member of the G-protein coupled chemokine receptor family that mediates hemopoietic cell proliferation, migration, homing, and cell adhesion to extracellular matrix molecules. CXCR4 is physiologically expressed on T and B lymphocytes, monocytes, macrophages, neutrophils, eosinophils, and hematopoietic stem cells in bone marrow [1]. Pathological CXCR4 overexpression has been reported in more than 30 various types of solid tumors and hematopoietic malignancies. It plays a crucial role in tumor growth, progression, invasiveness, cancer cellmicroenvironment interaction, and metastasis [2,3].
[ 68 Ga]pentixafor, a CXCR4-targeted PET probe with high affinity and selectivity to the receptor, has been developed and allows the sensitive and high-contrast PET imaging of CXCR4-expressing tissues and diseases in vivo [1]. The first clinical application of [ 68 Ga]pentixafor PET has been carried out in patients with non-Hodgkin lymphoma and multiple myeloma, which confirmed the CXCR4 expression in these lymphoproliferative diseases as a proofof-concept [4]. Since then, most studies of [ 68 Ga]pentixafor were focused on evaluation of hematologic malignancies, for example, multiple myeloma [1,[5][6][7], Waldenström macroglobulinemia/lymphoplasmacytic lymphoma [8,9], mucosa associated lymphoid tissue (MALT) lymphoma [10], chronic lymphocytic leukemia [11], and acute myeloid leukemia [12,13]. Some studies showed remarkable superiority of [ 68 Ga]pentixafor PET in detecting tumors and staging of the disease when compared with [ 18 F]FDG PET [5,7,8]. Besides the diagnostic use of [ 68 Ga]pentixafor PET, CXCR4-directed radioligand therapy with 90 Y-or 177 Lu-labeled Pentixather, the therapeutic partner of [ 68 Ga]pentixafor, has been successfully introduced as the compassionate use of treatment for relapsed, advanced stage multiple myeloma, diffuse large B cell lymphoma (DLBCL), and acute myeloid leukemia, in addition to high-dose chemotherapy regimens and followed by subsequent hematopoietic stem cell transplantation [13][14][15][16].
These studies depict the potential of CXCR4-directed PET imaging and radioligand therapy in hematologic malignancies. Considering the biologic differences of each tumor entity that may result in various levels of CXCR4 expression, it is necessary to study the CXCR4 expression with [ 68 Ga]pentixafor PET to help better select the patient for CXCR4-directed imaging and personalized therapy in future clinical applications. Herein, we performed this retrospective study to describe the [ 68 Ga]pentixafor PET/CT imaging in a spectrum of non-Hodgkin lymphomas and to compare it with [ 18 F]FDG PET/CT, which served as a reference. week. The imaging characteristics were analyzed, and quantitative parameters were measured retrospectively. The study was approved by the institutional review board of PUMCH (IRB protocol #ZS-1113), and written informed consent was obtained from each patient.

PET/CT imaging
The DOTA-CPCR4-2 peptide was purchased from CSBio Co (CA 94025, USA). The radiolabeling of [ 68 Ga]pentixafor was performed manually before injection according to the procedures as previously published [8]. [

Image analysis and statistics
All PET/CT scans were first visually rated in a binary fashion by 2 experienced nuclear medicine physicians (YL and QP), both of whom had over 7 years and 2 years experience in reading [ 18 F]FDG and [ 68 Ga]pentixafor PET/CT, respectively. Lesions were visually determined as focally increased tracer retention as compared to surrounding normal tissue. As previously described [7,8], bone marrow involvement in PET/CT was interpreted as being positive if there was presence of focal lesions with positive PET results, or diffuse bone marrow patterns with uptake higher than liver. The involvement of lymphoma and intensity of lesions were recorded. Visually, the tracer depicting higher number or intensity of tumor lesions was considered superior. Semi-quantitative analysis was performed with standard uptake value (SUV) and tumor-to-background ratios (TBR). A circular region of interest was placed over the tumor area with the maximal radioactivity, and SUVmax of the lesion was generated. Tumor uptakes were presented as SUVmax of the most avid lesion in a patient or in an organ. A reference blood pool region was defined by drawing a region of interest in the left atrium of the heart to derive TBR blood , and region of interest was drawn in liver to derive TBR liver .
Statistical analyses were done with the SPSS Statistics software (version 22.0, IBM SPSS Inc.). Comparison of numerical data of 2 groups was performed using Student's t test for data with normal distribution and Wilcoxon rank sum test for skewed data. A p value < 0.05 was considered statistically significant.

Results
The lymphomas of the 27 enrolled patients (19 male, 8 female; age, 57.2 ± 13.1 years, range 15-76 years) included lymphoplasmacytic lymphoma (n = 8), marginal zone lymphoma (n = 4), peripheral T cell lymphoma (n = 4), DLBCL (n = 3), unclassified indolent B cell lymphoma (n = 3), follicular lymphoma (n = 2), NK/T cell lymphoma (n = 2), and mantle cell lymphoma (n = 1). The patients did not receive any treatment against lymphoma before PET/CT scans. Clinical characteristics of the recruited patients and the diagnostic performance of dualtracer PET/CT in each case are given in Table 1, and examples of maximum intensity projections of the dualtracer PET scans in lymphomas are shown in Fig. 1 Table 2 and Fig. 2, respectively.

Lymphoplasmacytic lymphoma
Bone marrow is the predominant site of involvement in lymphoplasmacytic lymphoma, which was confirmed by bone marrow aspiration and biopsy in all recruited patients. On [ 68 Ga]pentixafor PET/CT, all 8 patients had intense radioactivity in the bone marrow, with an SUVmax of 9.5 ± 2.6 (range, 7.0-14.8 [ 68 Ga]pentixafor PET/CT detected positive lymph node involvements in 7 patients (SUVmax 11.3 ± 3.5, range 6.9-16.9), and 6 patients had involvement in more

Marginal zone lymphoma
Four patients were confirmed with marginal zone lymphoma at histology, including 3 patients with MALT    Lymphoplasmacytic lymphoma is an indolent non-Hodgkin lymphoma characterized by the accumulation of lymphoplasmacytic cells producing excessive monoclonal immunoglobulin in the bone marrow. As lymphoplasmacytic lymphoma is usually not avid for [ 18 F]FDG [17], [ 18 F]FDG PET/CT is not indicated for lymphoplasmacytic lymphoma unless there is a suspicion of aggressive transformation [18]. Consistent with our previous research [8,9], the current study again revealed an obvious superiority of [ 68 Ga]pentixafor to [ 18 F]FDG in staging of lymphoplasmacytic lymphoma. These results observed in clinical investigation were also supported by the fact that the CXCR4 expression is higher in the B cells of patients with lymphoplasmacytic lymphoma than in the B cells of healthy donors [19]. Therefore, the CXCR4-targeted PET/CT imaging with [ 68 Ga]pentixafor may become an important tool for diagnosis and staging of lymphoplasmacytic lymphoma.
The staging of marginal zone lymphoma is a challenge with [ 18 F]FDG PET/CT, because marginal zone lymphoma usually does not present with elevated glycolysis and may had heterogeneous metabolic behavior [20,21]. Stollberg S. and the colleagues found that CXCR4 expression was present at a high intensity in 92% of the cases [22]. In a recent study of MALT lymphoma, 33 [23]. In accordance with the above research, our study showed that the 4 patients with marginal zone lymphoma were detected by [ 68 Ga]pentixafor PET/CT with markedly increased radioactivity, which was superior or at least equal to [ 18 F]FDG PET/CT. This result indicates that [ 68 Ga]pentixafor PET/CT might be useful in detection and staging of marginal zone lymphoma; however, further study in larger patient cohort is warranted.
A previous study determined that CXCR4 was highly expressed in DLBCL cell lines, and in a patient with DLBCL, [ 68 Ga]pentixafor PET/CT resulted in excellent tumor uptake [4]. Due to the highly expressed CXCR4 in DLBCL, experimental CXCR4-directed radioligand therapy was also used as part of the conditioning regimen prior to allogeneic stem cell transplantation in several patients with relapsed advanced-stage DLBCL [16]. Our study added evidence to the overexpression of CXCR4 in DLBCL, although the uptake of [ 68 Ga]pentixafor was lower than the FDG uptake. For the lesions within the brain, [ 68 Ga]pentixafor exhibited a clear delineation with higher image contrast compared to [ 18 F]FDG due to the negligible uptake in cerebrum.
Strong CXCR4 expression is also detected in follicular lymphoma with both flow cytometry and immunohistochemical analysis [24,25]. We previously reported the findings of [ 68 Ga]pentixafor PET/CT in a patient with post-treated POEMS syndrome and concurrent follicular lymphoma, with active uptake of [ 68 Ga]pentixafor (SUVmax 9.7) [26]. The results of the current study corresponded well to the previous studies, and it suggested that [ 68 Ga]pentixafor might be useful in evaluation of follicular lymphoma in future.
Mantle cell lymphoma is usually an FDG-avid lymphoma, and [ 18 F]FDG PET/CT is a useful tool in staging and evaluating treatment response [27,28]. For CXCR4 expression, studies found that mantle cell lymphomas displayed high levels of CXCR4 expression, which is critical for malignant B cell trafficking and homing to supportive tissue microenvironment [25,29,30]. Consistent with these results, the only one patient with mantle cell lymphoma in our study had increased uptake of [ 68 Ga]pentixafor in the involved lymph nodes, despite the accumulation of [ 68 Ga]pentixafor that was obviously lower than that of [ 18 F]FDG. As there is no other clinical data on the presentation of [ 68 Ga]pentixafor PET in mantle cell lymphoma, it is hard to draw to a conclusion based on this single case. We think it might be interesting to further investigate the role of [ 68 Ga]pentixafor PET and the feasibility of CXCR4-directed radioligand therapy in mantle cell lymphoma.
There is no reported clinical data on the CXCR4targeted imaging in T cell and NK/T cell lymphomas, except a single case report of mycosis fungoides showing intense uptake of [ 68 Ga]pentixafor [31]. Weng AP and the colleagues reported that only 11.5% (3/26) of the cases with peripheral T cell lymphoma exhibited positive immunohistochemical staining for CXCR4 [32]. This low positive rate of CXCR4 expression explains the negative result of CXCR4-targeted PET imaging in the patient with PTCL-NOS in our study. Similarly, the two patients with NK/T cell lymphoma did not show increased uptake of [ 68 Ga]pentixafor in the tumors, which were very FDG-avid, suggesting the lack of CXCR4 expression in NK/T cell lymphoma as well.
EATL is a rare and aggressive subtype of extranodal T cell lymphoma arising in the gastrointestinal tract. The status of CXCR4 expression in EATL has not been reported yet. Our study has several limitations. First, the study has a small sample size. In some types of lymphoma, the number of cases was too small to perform a head-to-head comparison of the detection rate in lymphoma between [ 68 Ga]pentixafor and [ 18 F]FDG. Second, we only include several types of lymphoma in our study. In addition to the lymphomas included in the current study, chronic lymphocytic leukemia, acute lymphoblastic leukemia, and acute myeloid leukemia have been reported being positive on [ 68 Ga]pentixafor PET [4,[11][12][13]. In vitro studies with flow cytometry, reverse transcriptionpolymerase chain reaction, and immunohistochemical analysis also detected strong expression of CXCR4 in hairy cell leukemia, T cell lymphoblastic lymphoma/ leukemia, Sézary syndrome, angioimmunoblastic lymphoma, and anaplastic large cell lymphoma [32][33][34][35][36][37][38][39]. The CXCR4-targeted imaging with [ 68 Ga]pentixafor in these lymphoma/leukemia needs to be further investigated. Third, we did not perform the in vitro studies to confirm the expression of CXCR4 in the recruited cases. Previous studies have determined that [ 68 Ga]pentixafor binded with high specificity and selectivity to human CXCR4 and CXCR4 expression was correlated with cellular uptake of [ 68 Ga]pentixafor in lymphoma cell lines [4]. We think these results have provided strong evidence of the mechanism of CXCR4-mediated [ 68 Ga]pentixafor uptake in lymphoma. Finally, the heterogeneity of PET/CT protocols (e.g., uptake time, administered activity, use of 2 different PET/CT scanners, and reconstruction parameters) may bias the quantitative PET/CT measurements.

Conclusion
The CXCR4 expression varies in different types of non-Hodgkin lymphoma. Overexpression of CXCR4 was detected with [ 68 Ga]pentixafor PET/CT in lymphoplasmacytic lymphoma, marginal zone lymphoma, DLBCL, follicular lymphoma, mantle cell lymphoma, unclassified indolent B cell lymphoma, and EATL. However, PTCL-NOS and NK/T cell lymphoma may not present CXCR4 overexpression. When comparing with [ 18 F]FDG, [ 68 Ga]pentixafor showed higher uptake than [ 18 F]FDG did in lymphoplasmacytic lymphoma and marginal zone lymphoma.