Visual Abstract
Abstract
Desmoplastic small round cell tumor (DSRCT) is a rare, radiosensitive, yet difficult-to-treat sarcoma subtype affecting predominantly male adolescents. Extensive intraperitoneal seeding is common and requires multimodal management. With no standard therapy established, the prognosis remains poor, and new treatment options are needed. We demonstrate the clinical potential of C-X-C motif chemokine receptor 4 (CXCR4)–directed imaging and endoradiotherapy in DSRCT. Methods: Eight male patients underwent dual-tracer imaging with [18F]FDG and CXCR4-directed [68Ga]pentixafor PET/CT. A visual comparison of both tracers, along with uptake quantification in active DSRCT lesions, was performed. [68Ga]pentixafor uptake was correlated with immunohistochemical CXCR4 expression on tumor cells. Four patients with end-stage progressive disease underwent CXCR4-based endoradiotherapy. We report the safety, response by RECIST 1.1, and survival after endoradiotherapy. Results: Uptake of [68Ga]pentixafor in tumor lesions was demonstrated in all patients with DSRCT, providing diagnostic power comparable to [18F]FDG PET. Corresponding CXCR4 expression was confirmed by immunohistochemistry in all DSRCT biopsies. Finally, 4 patients were treated with CXCR4-directed [90Y]endoradiotherapy, 3 in a myeloablative dose range with subsequent autologous stem cell transplantation. All 3 required transfusions, and febrile neutropenia occurred in 2 patients (resulting in 1 death). Notably, severe nonhematologic adverse events were absent. We observed signs of response in all 3 patients, translating into disease stabilization in 2 patients for 143 and 176 d, respectively. In the third patient, postmortem autopsy confirmed a partial pathologic response. Conclusion: We validated CXCR4 as a diagnostic biomarker and a promising target for endoradiotherapy in DSRCT, demonstrated its feasibility, and provided the first evidence of its clinical efficacy.
Desmoplastic small round cell tumor (DSRCT) is an extremely rare malignant mesenchymal neoplasm that predominantly affects young men (1). The primary location is the abdominal cavity, in which is commonly found a multinodular disease affecting the omentum, retroperitoneum, and mesentery. Histologically, DSRCT is an aggressive sarcoma subtype that presents with multiphenotypic differentiation, including epithelial, muscular, and neural features, such as coexpression of cytokeratins, desmin, and synaptophysin. The recurrent balanced chromosomal translocation t(11:32)(p13;q12) is a pathognomonic hallmark and a driver of the disease. The corresponding EWSR1-WT1 fusion gene codes for a chimeric protein, with typically strong nuclear expression, containing the N-terminal domain of the Ewing sarcoma breakpoint region 1 protein and 3 of the 4 zinc finger domains of the Wilms tumor 1 protein (2,3).
DSRCT is characterized by immunologic ignorance (4). In particular, next-generation sequencing molecular profiling revealed a paucity of secondary mutations with notable heterogeneity between patients, and (except for FGFR4 mutations in only a small subset of patients), no suitable therapeutic targets could be identified (5).
Because of the lack of clinical trials in this orphan disease, with approximately 1,000 patients reported to date, no standard therapy has been established. Patients with DSRCT have been compiled in sarcoma studies, and systemic chemotherapy regimens are derived from protocols established primarily for Ewing and other soft-tissue sarcomas. Complete resection has been shown to increase overall survival (OS) (6); however, as primary curative surgery is rarely achievable, multimodal treatment with aggressive induction with or without high-dose chemotherapy with autologous stem cell transplantation followed by cytoreductive surgery with or without hyperthermic intraperitoneal chemotherapy (7) or whole-abdominal radiotherapy has been proposed (2,3). However, patients experience early relapse and prognosis remains poor, with a median OS of 24–29 mo and 3-y and 5-y survival rates of 30%–35% and 4%, respectively (8–10). So far, targeted therapies with sunitinib (11) and pazopanib (12,13), as well as immunotherapy with nivolumab (14) or pembrolizumab, have demonstrated only limited effects (15,16).
Functional imaging using [18F]FDG PET/CT is regarded as the most suitable imaging technique for DSRCT and helps to select patients with a metabolic response to induction chemotherapy for debulking surgery even in the absence of significant tumor shrinkage according to RECIST (17,18).
C-X-C motif chemokine receptor 4 (CXCR4) was first identified as a coreceptor for HIV (X4-tropic isolates) entry into cells. Beyond its role in various physiologic processes, including embryogenesis, angiogenesis, and modulation of hematopoietic stem cells (19,20), CXCR4 has gained attention because it is overexpressed in more than 20 different tumor entities (21–24), including sarcoma (25–27), with higher receptor expression denoting poor prognosis (23,28,29). In particular, among sarcomas, CXCR4 overexpression has been previously described in Ewing sarcoma, which shares many biologic features with DSRCT, providing a rationale for exploring and targeting CXCR4 in DSRCT (26). Recently, noninvasive visualization of CXCR4 in vivo using PET has become possible with the development of [68Ga]pentixafor (30). In addition, the first proof-of-concept studies have demonstrated the feasibility of subsequent receptor-directed endoradiotherapy in CXCR4-expressing diseases (21,31–33).
MATERIALS AND METHODS
Inclusion of Patients
This case study was approved by the Ethics Committee of the Medical Faculty, University of Würzburg (approval 20201001 01), and written informed consent was obtained from all patients before diagnostic and therapeutic procedures. [68Ga]pentixafor was offered in compliance with §37 of the Declaration of Helsinki and the German Medicinal Products Act Arzneimittelgesetz §13 2b.
Between October 2015 and April 2020, 8 young, male patients (median age at diagnosis, 29 y [range, 8–43 y]) with DSRCT (7 confirmed cases of DSRCT patients, 1 case of undifferentiated peritoneal small round cell sarcoma with clinical and morphologic features of DSRCT) underwent imaging with [18F]FDG and [68Ga]pentixafor PET/CT at our institution. At presentation, all patients had extensive disease, with Hayes–Jordan stage II (widespread intraabdominal lesions) in 2 patients, stage III (additional liver metastasis) in 1 patient, and stage IV (additional extraabdominal metastasis) in 5 patients (including 4 patients with liver metastasis; Table 1). Finally, 4 patients with advanced, unresectable, and progressive disease after conventional therapies were selected for CXCR4-directed [90Y]endoradiotherapy and after individual dosimetry received a total of 5 cycles of endoradiotherapy.
Imaging Protocol and Analysis
All patients underwent CXCR4-directed PET/CT ([68Ga]pentixafor, to noninvasively visualize the receptor expression of DSRCT lesions and evaluate eligibility for CXCR4-directed endoradiotherapy) and [18F]FDG PET/CT (as a control). Both PET/CT studies were performed on 2 consecutive days (the supplemental materials, available at http://jnm.snmjournals.org, provide a detailed description) (34). Briefly, PET/CT images were independently analyzed by 2 nuclear medicine specialists using a commercial software package (syngo.via, VB60A HF02; Siemens Healthineers AG). All lesions with nonphysiologic uptake of the respective tracer higher than the physiologic background were rated as CXCR4-positive or [18F]FDG-positive. For corresponding tumor lesions on [68Ga]pentixafor and [18F]FDG PET/CT, the average SUVmax within a spheric volume of 1 mL (SUVpeak) was recorded, and tumor-to-background ratios (TBRs) were calculated (details are provided in the supplemental materials) (35). For posttherapeutic tumor evaluation, [18F]FDG PET/CT imaging was performed. Tumor response was assessed by [18F]FDG PET/CT using RECIST 1.1 and PERCIST 1.0 (36,37).
Immunohistochemistry
In total, 9 biopsies from all 8 patients were examined for CXCR4 expression by immunohistochemistry (patient 2 had a second postmortem biopsy from a liver metastasis). The intensity of CXCR4 expression was visually rated using a 4-point scale (0, absent; 1, weak; 2, moderate; and 3, intense). The percentage of stained tumor cells was estimated, and the staining intensity was rated by the immunoreactive score (supplemental materials) (38).
CXCR4-Directed Therapy
For patients selected for CXCR4-directed endoradiotherapy, individual pretherapeutic dosimetry with [177Lu]pentixather, with a mean activity of 180 ± 45 MBq, was performed as previously described (39). Achievable doses in the tumor manifestations were estimated by multiplying the calculated dose coefficient in Gy/GBq by the administered activity of [90Y]pentixather. Standardized institutional protocols for the endoradiotherapy work-up were applied, as previously described (21,33). Drug-related adverse events and toxicities were evaluated according to the Common Toxicity Criteria of the National Cancer Institute (version 5.0) (40). Progression-free survival and OS were calculated from the date of endoradiotherapy until documented radiologic or clinical progression or death.
Statistical Analysis
Statistical analyses were performed using Prism, version 9.3.0 (GraphPad Software). Descriptive data are presented as mean ± SD or median and range. To test for a normal distribution, the Shapiro–Wilk test was applied and refuted a normal distribution in most of the imaging data (SUV and TBR). For statistical comparison of SUV and TBR for both tracers in corresponding lesions, as well as before and after therapy, the Wilcoxon signed-rank test was performed. P values of less than 0.05 were considered statistically significant.
RESULTS
Patient Cohort
The patient characteristics are detailed in Table 1 and the supplemental materials, and the treatment course is illustrated in Figure 1. Before CXCR4 imaging and subsequent endoradiotherapy (if the patient was eligible), all 8 patients received intensive multiagent induction chemotherapy according to established sarcoma protocols (e.g., EWING 2008 protocol) (41–43), followed by high-dose chemotherapy with autologous stem cell transplantation in analogy to Kushner et al. (3 patients) (44) or as conditioning chemotherapy along with endoradiotherapy (1 patient). A median of 2.5 lines (range, 1–5 lines) of previous systemic regimens were administered before CXCR4 imaging and subsequent endoradiotherapy. Seven patients underwent prior abdominal debulking surgery (6 with additional hyperthermic intraperitoneal chemotherapy), and 1 patient was subjected to radiation therapy of a single vertebral body before diagnostic CXCR4 PET. The median time from the start of first-line systemic therapy to endoradiotherapy was 15.1 mo (range, 7.3–33.4 mo), and the median interval between diagnostic CXCR4 PET and CXCR4-directed [90Y]endoradiotherapy was 48 d (range, 26–92 d). Notably, patient 2 received 2 subsequent cycles of endoradiotherapy.
Analysis of [68Ga]Pentixafor and [18F]FDG PET
All 7 patients with EWSR1-WT1 fusion-positive DSRCT showed a significant accumulation of [68Ga]pentixafor in tumor lesions, whereas patient 5, classified as having an undifferentiated sarcoma on reference pathology, was the only patient lacking tracer uptake. Although most of the tumor manifestations were concordantly CXCR4-positive and [18F]FDG-positive, we found discordant [18F]FDG-positive, CXCR4-negative manifestations in 3 patients (2 patients with only 1 lymph node metastasis and 1 patient with 3 peritoneal metastases). In contrast, discordant [18F]FDG-negative, CXCR4-positive manifestations were observed in 3 patients (2 patients with only 1 metastasis [liver in one and lymph node in the other] and 1 patient with 3 peritoneal metastases). Figure 2 shows an example of patient 2. Of an overall 61 tumor lesions, 60 showed [18F]FDG uptake (98.4%) and 57 showed [68Ga]pentixafor uptake (93.4%). Fifty-six lesions showed corresponding [18F]FDG and [68Ga]pentixafor uptake. Of these, the median SUVpeak (5.7 [range, 1.5–16.6] vs. 4.7 [range, 1.7–10.3], P ≤ 0.001) and median TBR (3.8 [range, 0.9–9.2] vs. 2.9 [range, 0.7–5.9], P ≤ 0.001) were significantly higher for [18F]FDG than for [68Ga]pentixafor.
Individual Dosimetry and Therapy with [90Y]Pentixather
After baseline screening with [68Ga]pentixafor PET/CT, all except 1 patient (patient 5) were considered eligible for CXCR4-directed endoradiotherapy. However, after interdisciplinary counseling, patient 4 opted for in-label treatment with pazopanib, whereas patients 6 and 8 had to be excluded because of compliance reasons and insufficient uptake in [177Lu]pentixather dosimetry, respectively. In the remaining patients, the mean estimated doses for tumor lesions were 2.7 ± 0.9 Gy/GBq (range, 1.4–3.6 Gy/GBq). Detailed dosimetry data are presented in Supplemental Table 1. Figure 3 shows an example of patient 7 before and after [90Y]pentixather endoradiotherapy.
Safety
Four patients received a total of 5 cycles of [90Y]pentixather with a mean activity of 6.6 ± 2.9 GBq (range, 1.7–9.1 GBq). Therapeutic administration of [90Y]pentixather was well tolerated, and no severe nonhematologic adverse effects (≥grade 3 Common Terminology Criteria for Adverse Events) occurred. On day 3, patients were transferred from the Department of Nuclear Medicine to Internal Medicine for further monitoring and autologous stem cell rescue on day 14 (after 5 half-lives of 90Y).
Severe hematotoxicity was expected and occurred in all patients with myeloablative endoradiotherapy, resulting in grade 3–4 neutropenia (febrile neutropenia in 2 patients, onset from days 10–12), grade 4 thrombocytopenia (onset from days 10–18) requiring an average of 3 platelet concentrates, and prolonged grade 3–4 anemia requiring an average of 3 red blood cell units per patient to bridge the time until blood count recovery in all patients (Table 2). One patient with end-stage disease and preexisting obstructive jaundice from extensive liver metastases died of endoradiotherapy-induced neutropenia after the second treatment cycle from septic cholangitis on the day of planned stem cell rescue.
Antitumor Efficacy of Endoradiotherapy with [90Y]Pentixather
Follow-up [18F]FDG PET/CT demonstrated a significant decrease in the median SUVpeak after therapy (4.7 [range, 2.2–14.4] vs. 7.4 [range, 1.9–16.6], P ≤ 0.001). In parallel, the median TBR also significantly declined (3.6 [range, 1.7–6.5] vs. 5.0 [range, 1.3–9.2], P ≤ 0.001).
All 3 patients treated with myeloablative activity had signs of metabolic response, and 2 (patients 3 and 7) achieved stable disease according to RECIST. The third patient (patient 2) demonstrated a metabolic response in preexisting lesions but was classified as having PERCIST progressive disease because of new [18F]FDG-positive lesions on the first follow-up imaging. This very frail patient with end-stage progressive disease and obstructive jaundice due to extensive liver metastases demonstrated indirect evidence of a response, with a temporary improvement in serum biochemistry, namely a 50% reduction in peak bilirubin levels after the first cycle of endoradiotherapy. Therefore, he continued to a second endoradiotherapy cycle but died from neutropenic sepsis 15 d after the second [90Y]pentixather application, with no additional systemic chemotherapy applied. Notably, in this patient, evidence of regression in 30%–50% of tumor cells (fulfilling the pathologic partial-response criteria) was demonstrated on autopsy (Supplemental Fig. 1). Interestingly, the only patient in our endoradiotherapy cohort with no metabolic response had received a significantly reduced, nonmyeloablative [90Y]pentixather activity because of lack of an autologous stem cell graft.
In summary, the progression-free survival of the cohort after endoradiotherapy was 104 d (range, 28–176 d), with the 2 evaluable patients demonstrating a promising progression-free survival of 143 and 176 d, respectively. Median OS of the total cohort from the start of first-line CT was 24.6 mo (range, 12.1–42.8 mo). This compares with survival data for DSRCT cohorts published in the literature, with OS varying between 24 and 29 mo (8–10).
Detailed information is summarized in Table 3 and the supplemental materials.
CXCR4 Immunohistochemistry
Moderate to strong membranous CXCR4 expression was detected by immunohistochemistry in 7 of 8 patients, with 55%–95% (median, 80%) positive cells. The immunoreactive score was predominantly in the middle range (6–8 points). Proliferative activity ranged from 20% to 70% (Ki-67), without association with CXCR4 labeling indices (Supplemental Table 2).
Patient 8 showed only 55% CXCR4-positive tumor cells, a finding that was associated with insufficient uptake during dosimetry. Patient 5, with morphologic and histologic features of DSRCT but no expression of EWSR1-WT1 fusion transcript, was classified as having undifferentiated sarcoma by reference pathology. This was the only patient completely negative for CXCR4 on immunohistochemistry and [68Ga]pentixafor PET/CT. In patient 2, CXCR4 expression was also examined in the liver metastasis at autopsy after CXCR4 endoradiotherapy. Interestingly, CXCR4 expression level was still high (primary biopsy, 95%; autopsy material, 80%), and morphologically distinct signs of regression were present, with increased cell and nuclear pleomorphism as well as tumor necrosis (Supplemental Fig. 1).
DISCUSSION
Here, we report a cohort of 8 male patients with DSRCT who underwent imaging with [18F]FDG and subsequent [68Ga]pentixafor PET/CT as screening for potential CXCR4-directed endoradiotherapy. CXCR4 expression has been previously described in different types of sarcoma, especially Ewing sarcoma, which shares many biologic features with DSRCT (25–27).
Radiation therapy is effective for DSRCT and has been shown to improve OS in some patients (9,45,46). However, whole-abdomen radiotherapy is associated with considerable toxicity in several organs at risk, resulting in significant dose reductions, and its benefits remain controversial. By leveraging the radiosensitive properties of DSRCT, we hypothesized that delivering radiotherapy on the molecular or cellular level targeting CXCR4 might reduce toxicity and offer a new treatment approach. CXCR4 endoradiotherapy has been shown to be safe and effective for different hematologic malignancies (9,45).
In our case series, we are the first—to our knowledge—to describe CXCR4 expression (confirmed by immunohistochemistry) in all cases of EWSR1-WT1 fusion-positive DSRCT. Notably, all these patients showed significant uptake of [68Ga]pentixafor in their tumor lesions.
Comparative imaging with [18F]FDG and [68Ga]pentixafor PET/CT showed comparable detection rates of 98.4% for [18F]FDG and 93.4% for [68Ga]pentixafor PET/CT between the tracers, with sensitivity levels comparable to previously published data (18). Lesions with exclusive CXCR4 positivity or negativity likely represent tumor heterogeneity and may respond differently to treatment. These lesions should be monitored and further investigated by targeted biopsies and may offer potential for individualized treatment decisions, such as targeted irradiation or surgery.
Four of our patients were treated with CXCR4-directed [90Y]endoradiotherapy after individual pretherapeutic dosimetry with [177Lu]pentixather. In accordance with previous experience with CXCR4-directed endoradiotherapy in other entities (21,31–33,47), myelosuppression due to CXCR4 expression on hematopoietic cells requires obligatory stem cell rescue, blood product support, and management of febrile neutropenia. This expected hematotoxicity is manageable in a generally young, usually organ-fit DSRCT patient population.
Interestingly, metabolic activity was significantly decreased in tumor lesions after endoradiotherapy, as measured by [18F]FDG PET/CT, indicating a therapeutic response (17). On a patient basis, 3 of 4 patients demonstrated signs of metabolic response, with the only nonresponder being probably underdosed (because of lack of stem cells). Remarkably, 2 patients with progressive disease before endoradiotherapy achieved disease stabilization with an OS of 225 and 282 d, respectively, and in the third patient, with fatal sepsis, a pathologic partial response was demonstrated on postmortem biopsy.
Promising results from an intraperitoneally applied radioligand (131I-omburtamab) have been reported by others in DSRCT, with a superior OS of 54 mo as compared with historical data from the standard of care at 36 mo (48). However, this local therapy differs from our systemic treatment in 3 ways: only patients after cytoreductive surgery (i.e., without a measurable disease burden or with a low disease burden) were included, intraperitoneal radioimmunotherapy does not target extraabdominal disease, and almost all patients received additional whole-abdomen radiotherapy; thus, the effect of intraperitoneal radioimmunotherapy alone remains unclear. Supporting data were recently published, targeting fibroblast activation protein with 90Y-labeled fibroblast activation protein inhibitor 46 radioligand therapy in a cohort that included 16 patients with advanced sarcoma (but no DSRCT). This approach was demonstrated to be safe and induced disease stabilization (RECIST 1.1) and metabolic responses (PERCIST) in approximately one third of the patients (49).
In our cohort, all treated patients presented with late-stage therapy-refractory DSRCT. We believe that more pronounced responses might be achievable with less tumor burden or an earlier disease stage. Given the descriptive and exploratory character of our analysis, we want to emphasize that our results have to be interpreted with caution and that the value of statistical analyses is severely compromised by the limited number of patients. However, considering the poor OS rates in DSRCT and the lack of standardized treatment, the medical need for innovative therapies is obvious. Thus, our proof-of-concept study could serve as a stimulus for future research and clinical trial design. For instance, we propose to investigate CXCR4 endoradiotherapy in DSRCT after early cytoreductive surgery or as part of a consolidating high-dose chemotherapy approach in patients with chemosensitive disease. In addition, modulation of CXCR4 receptor expression, as recently described by others (50–52), needs to be explored for its potential to increase endoradiotherapy efficacy. Finally, our promising data clearly indicate that CXCR4-directed endoradiotherapy may also be exploited in Ewing sarcoma, which occurs much more frequently and is known to overexpress CXCR4 (26,53).
CONCLUSION
CXCR4 is a promising diagnostic and therapeutic biomarker for DSRCT, as confirmed by immunohistochemistry and PET. We demonstrated the feasibility of CXCR4-directed endoradiotherapy and provided the first evidence of its antitumor activity.
DISCLOSURE
This work was supported by the IZKF Wuerzburg (grant Z-02/85 to Philipp Hartrampf). Hans-Jürgen Wester is a founder and shareholder of Scintomics. No other potential conflict of interest relevant to this article was reported.
KEY POINTS
QUESTION: Can CXCR4-directed endoradiotherapy be performed on DSRCT, a radiosensitive, yet difficult-to-treat sarcoma?
PERTINENT FINDINGS: Since DSRCT overexpresses CXCR4 on the cell surface, receptor-directed PET imaging and subsequent endoradiotherapy are feasible. Beyond the expected hematotoxicity, CXCR4-directed endoradiotherapy was well tolerated, with no severe nonhematologic adverse events, and was able to induce disease stabilization in patients with advanced DSRCT.
IMPLICATIONS FOR PATIENT CARE: CXCR4-directed endoradiotherapy in DSRCT is feasible and might prove a new option for patients with otherwise limited treatment alternatives.
Footnotes
Published online Jun. 22, 2023.
- © 2023 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication January 17, 2023.
- Revision received April 14, 2023.