Chemokine receptor–targeted PET/CT provides superior diagnostic performance in newly diagnosed marginal zone lymphoma patients: a head-to-head comparison with [¹⁸F]FDG

Representing approximately 7% of all indolent non-Hodgkin’s lymphomas, marginal zone lymphomas (MZL) comprised three distinct subtypes: extranodal MZL (EMZL) of mucosa-associated lymphoid tissue, nodal MZL (NMZL), and splenic MZL (SMZL) [1, 2]. For initial workup of newly diagnosed MZL, current practice guidelines recommend contrast-enhanced computed tomography (CT) as imaging modality of choice [3]. The value of 2-deoxy-2-[¹⁸F]fluoro-D-glucose ([¹⁸F]FDG) positron emission tomography/CT (PET/CT) for routine initial staging is a matter of debate, with reported variable radiotracer accumulation depending on MZL subtype, or other histopathologic and morphologic features [4‐9]. Nonetheless, respective guidelines recommend the use of this PET agent in challenging scenarios, e.g., for guidance of biopsies or treatment monitoring of localized treatment [3].

Beyond [¹⁸F]FDG assessing glucose consumption in sites of disease, PET agents specifically targeting other (sub)cellular structures overexpressed in MZL have been evaluated in recent years [10, 11]. For instance, the C-X-C motif chemokine receptor 4 (CXCR4)-directed radiotracer [⁶⁸Ga]Ga-PentixaFor showed improved read-out capabilities when compared to guideline-compatible work-up, including CT. Those previous results, however, did not focus on a comprehensive head-to-head comparison of [⁶⁸Ga]Ga-PentixaFor vs [¹⁸F]FDG in all lymphoma manifestations. As such, we aimed to compare findings on both PET/CTs in untreated MZL patients, including a visual and quantitative assessment on a patient and target lesion (TL)-level specifically focusing on the whole body lymphoma load.

We performed a retrospective analysis of our institutional PET/CT database and identified 32 newly diagnosed MZL patients who received [⁶⁸Ga]Ga-PentixaFor and [¹⁸F]FDG PET/CT within no more than 30 days. Patients did not receive oncological treatment between PET scans. All patients provided written informed consent. The local institutional review board waived the need for additional approval as this was a retrospective investigation (waiver no. 20220414 01). Prior studies have reported on parts of this patient cohort [10, 12, 13]. In [10], only a “hottest lesion” analysis was conducted; i.e., exclusively, the most intense lesion on PET has been further investigated, while in the present study, a lesion-based head-to-head comparison of the entire lymphoma burden has been performed.

Mean patient age was 65.6 ± 13.4 years. 19/32 (59.4%) patients were female, and 13/32 (40.6%) were male. Distribution among subtypes based on molecular imaging findings was as follows: 17/32 (53.1%) NMZL, 13/32 (40.6%) EMZL, and 2/32 (6.3%) SMZL.

In the present visual and quantitative analysis of the total lymphoma load on [¹⁸F]FDG and [⁶⁸Ga]Ga-PentixaFor PET/CT, we observed that the latter radiotracer revealed MZL manifestations in a higher proportion of patients, along with substantially increased detection rate on a TL level. In a read-out of quantitative PET parameters for concordant lesions on both scans, [⁶⁸Ga]Ga-PentixaFor almost consistently provided significantly higher SUV_max, SUV_peak, and TBR, indicative for improved contrast. In a subgroup analysis of MZL subtypes, visual and quantitative comparison of both radiotracers also provided higher diagnostic performance of [⁶⁸Ga]Ga-PentixaFor PET/CT for NMZL and EMZL. As such, based on our visual and quantitative comparison, the latter PET agent may emerge as the radiotracer of choice when investigating MZL patients upon initial diagnostic work-up.

As the reference radiotracer in nuclear oncology, [¹⁸F]FDG is well-established in the vast majority of lymphomas, in particular for treatment monitoring [15]. Not surprisingly, this agent has also been extensively validated in patients with MZL, but its diagnostic accuracy seems to vary based on the investigated subtype. For instance, Hofmann et al. reported that [¹⁸F]FDG PET is positive in nodal, but not EMZL, which is in line with our findings reporting on limited detection rate for the latter subtype on a TL level (Table 2) [16]. Nonetheless, the clinical value of this radiotracer remains a matter of debate, as a recent study investigating all three MZL subtypes reported on consistent upstaging after [¹⁸F]FDG administration when compared to CT alone [17]. Given those controversial findings, novel radiotracers providing a more reliable read-out of the current disease extent are intensively sought. In this regard, theranostic radiotracers may provide improved staging along with potential therapeutic options using ß-emitting 177Lu-labelled counterparts. Recent immunohistochemical analysis, however, revealed limited or even abundant expression of somatostatin receptors (SSTR) 2a/5 [18, 19], which are specifically addressed by SSTR-directed radiotracers used for imaging and therapy [20]. Not surprisingly, in vivo targeting of this receptor subtype then provided a substantial rate of false-negative findings, thereby rendering SSTR as less relevant in the context of MZL [21]. CXCR4, however, was substantially upregulated in more than 90% of samples obtained from extranodal MZL of mucosa-associated lymphoid tissue lymphoma [18]. Also exploited as theranostic target, such chemokine receptors have been subject to multiple imaging studies using [⁶⁸Ga]Ga-PentixaFor [10‐12]. For instance, a recent analysis focusing on the diagnostic performance of CXCR4-directed PET/CT relative to guideline-compatible diagnostic work-up (CT, bone marrow biopsy and esophagogastroduodenoscopy) also provided preliminary evidence on the superiority of this radiotracer relative to [¹⁸F]FDG. However, in this previous study, only one single TL, defined as most intense on uptake, was analyzed [10]. In the present investigation, we expanded those preceding findings by examining the entire whole-body lymphoma burden on a visual and quantitative level, which then also provided a comprehensive compartment- and subtype-based analyses. Of note, for both lymphonodal and extranodal compartments, CXCR4-targeted imaging revealed substantially more sites of disease which would have been missed by [¹⁸F]FDG, in particular in manifestations located in the soft tissue and bone. Although limited by the retrospective nature and small number of enrolled subjects, our study may provide a hint that [⁶⁸Ga]Ga-PentixaFor may emerge as the novel diagnostic PET agent of choice in MZL. These considerations are also further fueled by an observed elevated TBR in both NMZL and EMZL. Indicative for improved image contrast, those findings suggest that CXCR4-directed imaging may address the clinical need of a reliable diagnostic PET agent, in particular in EMZL, where [¹⁸F]FDG only provides rather limited clinical benefit. In this regard, future studies should determine whether the observed high image contrast on [⁶⁸Ga]Ga-PentixaFor PET indeed improve reader’s confidence, e.g., by conducting an interobserver agreement study [22]. Nonetheless, chemokine receptor imaging may be less suitable in patients with splenic involvement, which is partially explained by the physiological biodistribution of this agent [23].

Beyond image contrast, we also recorded SUV_max of median >10, with individuals even exhibiting values above 37. Based on those PET metrics, therapeutic applications using [¹⁷⁷Lu]Lu- or [⁹⁰Y]Y-PentixaTher would then provide anti-lymphoma efficacy, along with (desired) eradication of the stem cell niche as an integral component of the treatment plan, e.g., to prepare for subsequent stem cell transplantation [24]. As such, relative to [¹⁸F]FDG, the theranostic aspect of CXCR4-targeted imaging would then not only allow for better delineation of putative sites of disease, but also identify patients eligible for treatment.

Last, several limitations of this study should be noted. First, this study’s retrospective nature implies, that it is not possible to histologically confirm specific lesions seen on molecular imaging. In addition, not all lesions that were conspicuous on imaging can be biopsied and confirmed histologically. However, the fact that several readers verified inconclusive lesions in consensus should contribute to diagnostic confidence. Moreover, previous work has shown that CXCR4-directed imaging can be used as an accurate staging tool in patients with marginal zone lymphoma [10]. In this regard, Duell and colleagues [10] demonstrated that 16 out of 18 PET-guided biopsies of suspicious lesions exclusively visualized on [⁶⁸Ga]Ga-PentixaFor PET compared to [¹⁸F]FDG were actually histologically confirmed as MZL manifestations. Therefore, even lacking bioptic confirmation of each lesion documented on imaging, we believe that viewed with due caution, our results are of value.

Compared to [¹⁸F]FDG, [⁶⁸Ga]Ga-PentixaFor PET/CT identified more sites of disease in a higher proportion of patients with newly diagnosed MZL, irrespective of the investigated subtype. PET quantification — including TBR — provided consistently higher values on [⁶⁸Ga]Ga-PentixaFor PET, thereby also suggesting improved image contrast. As such, in patients affected with MZL, our findings may pave the way for a more widespread adoption of this PET agent in the clinic.