Lymphoma-Sink Effect in Marginal Zone Lymphoma Based on CXCR4-Targeted Molecular Imaging

Purpose Recent studies investigating a tumor-sink effect in solid tumors reported on decreasing uptake in normal organs in patients with higher tumor burden. This phenomenon, however, has not been evaluated yet for theranostic radiotracers applied to hematological neoplasms. As such, we aimed to determine a potential “lymphoma-sink effect” in patients with marginal zone lymphoma (MZL) imaged with C-X-C motif chemokine receptor (CXCR) 4-directed PET/CTs. Procedures We retrospectively analyzed 73 patients with MZL who underwent CXCR4-directed [68Ga]Ga-PentixaFor PET/CT. Normal unaffected organ uptake (heart, liver, spleen, bone marrow, kidneys) was quantified using volumes of interests (VOIs) and mean standardized uptake values (SUVmean) were derived. MZL manifestations were also segmented to determine the maximum and peak standardized uptake values SUV (SUVmax/peak) and volumetric parameters, including lymphoma volume (LV), and fractional lymphoma activity (FLA, defined as LV*SUVmean of lymphoma burden). This approach resulted in 666 VOIs to capture the entire MZL manifestation load. We used Spearman’s rank correlations to determine associations between organ uptake and CXCR4-expressing lymphoma lesions. Results We recorded the following median SUVmean in normal organs: heart, 1.82 (range, 0.78–4.11); liver, 1.35 (range, 0.72–2.99); bone marrow, 2.36 (range, 1.12–4.83); kidneys, 3.04 (range, 2.01–6.37); spleen, 5.79 (range, 2.07–10.5). No relevant associations between organ radiotracer uptake and MZL manifestation were observed, neither for SUVmax (ρ ≤ 0.21, P ≥ 0.07), SUVpeak (ρ ≤ 0.20, P ≥ 0.09), LV (ρ ≤ 0.13, P ≥ 0.27), nor FLA (ρ ≤ 0.15, P ≥ 0.33). Conclusions Investigating a lymphoma-sink effect in patients with hematological neoplasms, we observed no relevant associations between lymphoma burden and uptake in normal organs. Those observations may have therapeutic implications, e.g., for “cold” SDF1-pathway disrupting or “hot,” CXCR4-directed radiolabeled drugs, as with higher lymphoma load, normal organ uptake seems to remain stable. Supplementary Information The online version contains supplementary material available at 10.1007/s11307-023-01830-9.


Introduction
C-X-C motif chemokine receptor 4 and its only known ligand SDF1 are crucially involved in stem cell mobilization and migration of CXCR4-positive bone marrow (BM) cells by following a gradient towards (neo-)angiogenic niches, finally Johannes Duell and Rudolf A. Werner contributed equally to this work.
leading to promotion of tumor-feeding vessels and metastatic spread [1]. For instance, in specimen of patients afflicted with marginal zone lymphoma (MZL), more than 90% revealed relevant CXCR4 expression and those ex-vivo findings then laid the proper groundwork for in-vivo imaging using CXCR4directed [ 68 Ga]Ga-PentixaFor PET/CT [2,3]. Further favoring a more wide-spread adoption in hemato-oncology, hematological neoplasms achieved substantially higher target-to-background ratios and maximum standardized uptake values (SUV max ) on a quantitative level when compared to most solid tumor entities [4]. As such, beyond enabling a more precise diagnostic readout, CXCR4-targeted PET/CT may also allow to identify candidates for treatment with ß -(minus)-emitting counterparts, e.g., [ 177 Lu]Lu-/[ 90 Y]Y-PentixaTher [5]. Of note, such a molecular imaging-based treatment strategy may then achieve both antilymphoma and desired myeloablative effects, thereby preparing for subsequent hematopoietic stem cell transplantation (HSCT) in addition to conditioning regimens [6,7].
Increasing tumor load, however, may be linked to decreasing radiotracer accumulation in normal organs, possibly explained by a lower amount of tracer available to the normal organs in the presence of high tumor load. Such a tumor-sink effect has been described for solid tumors including prostate carcinoma (PC) imaged with prostate-specific membrane antigen (PSMA)targeted PET [8]. A recent evaluation using chemokine receptor imaging in patients diagnosed with other solid tumor entities than PC did not reveal such an interdependence on an organ-tumor level [9]. For hematological neoplasms, a potential "lymphomasink" effect has been shown in patients with diffuse large B cell lymphoma (DLBCL) imaged with 2-[ 18 F]FDG PET/CT [10,11]. For CXCR4-targeted PET/CT, however, a lymphoma-sink effect has not been determined yet. In a theranostic setting, those information may be of relevance, as in patients with increasing lymphoma load, higher [ 177 Lu]Lu-/[ 90 Y]Y-PentixaTher activities could then be administered, thereby optimizing anti-lymphoma efficacy and minimizing side effects in organs at risk [6,12]. As such, investigating the largest cohort of MZL patients imaged with [ 68 Ga]Ga-PentixaFor to date, we aimed to investigate whether such a potential lymphoma-sink effect is also present in individuals with MZL scheduled for chemokine receptor PET.

Methods
We retrospectively identified 73 patients with MZL who underwent [ 68 Ga]Ga-PentixaFor PET/CT from our institutional PET/CT database. The local institutional review board waived the need for further approval due to the retrospective nature of this study (waiver no° 20210726 02). Written informed consent of all patients was collected beforehand. Parts of this patient cohort have been investigated in previous studies [3,4], however, without examining potential implications of lymphoma burden on normal organ radiotracer uptake.

Imaging Procedure
Biograph mCT64 or 128 (Siemens Healthineers, Erlangen, Germany) PET/CTs were used for [ 68 Ga]Ga-PentixaFor exams. The scans raged from vertex of the skull to the proximal thighs. A mean of 128 MBq (± 24.5 MBq) of [ 68 Ga]Ga-PentixaFor was applied and images were acquired approximately 1 h after injection. PET/CT scans were acquired in 3D mode with an acquisition time of 2 min/bed position (mCT64) or continuous bed motion with flow bed velocity of 1.1 mm/s (mCT128). For image reconstruction, Gaussian filter was set to 2.0 mm using a matrix of 200*200 at 3 iterations with 24/21 subsets (mCT 64/mCT 128), using point spread function/time of flight (mCT 64/mCT 128). CT scans with or without contrast enhancement were used for attenuation correction [3,9].

Image Analysis
Images were analyzed on a dedicated workstation using syngo.via software (version VB60A; Siemens Healthineers, Erlangen, Germany). [ 68 Ga]Ga-PentixaFor normal organ uptake was established by one reader (S.Sc.) by placing spherical volumes of interest (VOI) with a minimum diameter of 10 mm as previously described [9,[13][14][15]. To determine mean standardized uptake values (SUV mean ) for normal unaffected organs, in each patient a sum of 8 VOIs in the liver, spleen, lateral myocardial wall of the left ventricle, both kidneys, and the vertebral bodies of C2, Th7, and L5 was drawn [9,[13][14][15]. For mean kidney and BM uptake, the average of the respective organ's VOIs was considered. To account for potential bone marrow involvement [16], the bone marrow was considered an unaffected normal organ only in patients with negative bone marrow biopsy results within 30 days of CXCR4-directed PET/CT. In addition, VOIs were also placed completely covering each MZL manifestation showing radiotracer accumulation above background levels, which automatically generated a 3-dimensional VOI at a 40% isocontour threshold. Inconclusive findings were verified by expert readers (S.Se., R.A.W., A.K.). Different quantitative aspects of total MZL load were derived, including averaged maximum (SUV max ), peak (SUV peak , defined as average SUV within a 1cm 3 sphere around the hottest voxel) and SUV mean , sum of lymphoma volume (LV in cm 3 ), and sum of fractional lymphoma activity (FLA, defined as lymphoma SUV mean *LV) [9,[13][14][15]. MZL manifestations smaller than 15 mm or 1.7 cm 3 were not sampled to account for potential partial volume effects [13,14,17].

Statistical Analysis
GraphPad Prism version 9.3.1 (GraphPad Prism Software, La Jolla, CA, USA) was utilized for statistical analyses. Gaussian distribution was determined using the Shapiro-Wilk test. We used Spearman's rank correlation coefficient (ρ) for correlative analyses [9,13]. P < 0.05 was considered statistically significant.

Correlative Analyses of Radiotracer Uptake and Lymphoma Manifestation Load
A comprehensive summary of our correlative analyses is provided in Table 3.

Discussion
Investigating a potential lymphoma-sink effect on CXCR4targeted PET/CT for hematological malignancies, we evaluated a cohort of individuals diagnosed with MZL, which presented with a broad range of lymphoma load and in-vivo tracer uptake on PET. We did not observe relevant associations between normal organ uptake and lymphoma burden, thereby indicating that such a phenomenon may rather not be evident in MZL patients imaged with [ 68 Ga]Ga-PentixaFor PET. Those findings may be of relevance, as recent reports provided evidence that CXCR4-targeted theranostics is particularly useful in patients with advanced blood cancers, especially for MZL [3,4]. In this regard, CXCR4-targeted radioligand therapy (RLT) based on pretherapeutic PET/CT has already achieved outcome benefits in end-stage lymphoma patients [5,18,19], thereby rendering this theranostic concept favorable for the referring hemato-oncologist [6]. Based on our findings, dose to unaffected organs may not decrease in patients with increasing CXCR4-positive lymphoma load or with elevated PET signal. Thus, other aspects, such as intra-or interpatient variability of uptake in normal  organs may be of more importance to consider dosimetry for CXCR4-RLT planning than the lymphoma load [15,20]. Recent reports investigating a tumor-sink effect in the context of theranostic radiotracers provided dissimilar results, including PC imaged with 18F-or 68Ga-labeled PET radiotracers or in patients scheduled for fibroblast activation protein-inhibitor-or somatostatin receptor (SSTR)directed imaging [8,[13][14][15]21]. For instance, our group Table 3 Spearman correlation (ρ) between normal organ radiotracer uptake and total MZL manifestation maximum and peak standardized uptake value (SUV max and SUV peak ), lymphoma volume (LV, in cm 3 ), and fractional lymphoma activity (mean SUV*LV). A trend towards significance was only seen for myocardial uptake with lymphoma-derived SUV max/peak Ga-PentixaFor administration and reported on no impact of tumor burden uptake on normal organ uptake [9]. Those studies are of importance, as all of those radiotracers can be used in a theranostic approach, i.e., after having determined the presence of the target, treatment with ß -(minus)emitting therapeutic equivalents can be conducted [22]. In this regard, higher activities could be administered for radiotracers harboring such a tumor-sink effect, which would then potentially result in higher doses to the tumor, but not to unaffected organs [21]. To date, however, investigations on a tumor-sink effect determined from baseline PET using theranostic tracers have been restricted to patients with solid tumors, including PC, neuroendocrine neoplasms, pancreatic cancer, lung or adrenocortical carcinoma [8,9,[13][14][15]21]. Recent years, however, have witnessed an increasing evaluation of CXCR4-directed imaging and therapy in particular for patients with hematological malignancies, thereby expanding this theranostic concept towards hematooncology [6]. While a potential lymphoma-sink effect has been observed on 2-[ 18 F]FDG PET/CT in patients with DLBCL [10,11], dedicated studies investigating a potential lymphoma-sink effect using the theranostic PET probe [ 68 Ga]Ga-PentixaFor are still missing. To date, RLTs with the therapeutic equivalent [ 177 Lu]Lu-or [ 90 Y]Y-PentixaFor have been conducted based on pretherapeutic dosimetry to determine the appropriate amount of activity [5,12,19,23,24]. Based on our results, however, image findings on [ 68 Ga]Ga-PentixaFor PET may not justify exceptional high treatment activities. Thus, while CXCR4-directed PET is required to determine the presence of the target in-vivo [6], those scans can rather not replace dosimetry prior to RLT. In hemato-oncology, recent efforts also turned towards "cold" CXCR4 inhibitory therapies, e.g., by investigating CXCR4 antagonist IgG1 antibodies, which achieved substantial anti-lymphoma effects by disrupting SDF1 pathways [25]. Preclinical models showed relevant benefit in multiple myeloma, acute myeloid leukemia or Hodgkin Lymphoma [25]. Of note, in all of those malignancies, [ 68 Ga]Ga-Pentix-aFor has provided evidence on visualizing chemokine receptor expression in sites of disease [7,26,27]. As such, given those recent developments of CXCR4-directed, non-radiolabeled drugs [25], our findings on a missing lymphoma-sink effect on PET may be of relevance not only for "hot," but also for "cold" CXCR4 inhibitory therapies, including dosing studies.
This investigation has limitations, e.g., its retrospective design and small number of subjects. Also, no additional weighting was made for different volumes of segmented lymphoma lesions when calculating SUV values within the same patient. Additionally, a potential lymphoma sink effect may be underestimated because only a small number of patients in our study had very high LV or FLA. Moreover, due to the retrospective nature of this study, we were not able to provide data on an ex-vivo blood count analysis, which might have offered further insights on a potential lymphoma sink effect by measuring blood pool activity. Furthermore, while previous investigations focusing on other theranostic radiotracers have mainly applied relative thresholds for tumor delineation [13,15], a recent investigation also used a fixed intra-individual threshold [8], e.g., based on the liver uptake. As this approach might allow for a more comparable tumor load delineation between scans, future studies may also conduct a respective re-analysis, preferably in a larger cohort with a broader range of LV. Last, we focused on MZL, as relative to others, this type of hematological malignancy has been well characterized by CXCR4-directed imaging including superiority to standard diagnostic work-up [3,4], indicative for a more widespread use in patients with MZL in the near-term future. However, [ 68 Ga]Ga-PentixaFor has also provided reliable information on the current chemokine receptor status in other hematooncological entities [4]. As such, future studies may also evaluate a potential lymphoma-sink effect in other subtypes, in particular in those which have already benefitted from CXCR4-RLT, including T-cell lymphoma [5].

Conclusions
We observed no relevant associations between normal organ uptake and CXCR4-positive lymphoma burden in patients with MZL studied with [ 68 Ga]Ga-PentixaFor PET/CT. As such, relative to other theranostic radiotracers used for imaging and therapy of solid tumors, a lymphoma-sink effect may rather not be evident in MZL. Those findings may be of relevance in a theranostic setting, as recent reports provided evidence that CXCR4-targeted RLT is particularly useful in patients with advanced blood cancers. In addition, the herein described missing lymphoma-sink effect may also be of relevance for dosing studies using novel CXCR4-inhibitory "cold" drugs.
Funding Open Access funding enabled and organized by Projekt DEAL. Okayama University "RECTOR" Program, KAKENHI grant (22H03027) from the Japan Society for the Promotion of Science (TH) and German Research Foundation (453989101, RAW, TH; 507803309, RAW).

Data Availability
The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations
Ethics Approval This study was conducted according to the guidelines of the Declaration of Helsinki. The need for further ethical review and approval were waived by the local Ethics Committee due to the retrospective character of the study (waiver no. 20210726 02).

Informed Consent
Procedures were performed as part of clinical care. Informed consent was obtained from all subjects.
Conflict of Interest RAW and AKB have received speaker's honoraria from PentixaPharm. AKB is a member of the advisory board of Pen-tixaPharm. JD has received honoraria from Incyte, Morphosys and Stemline. All other authors declare no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.