99mTc-CXCR4-L for Imaging of the Chemokine-4 Receptor Associated with Brain Tumor Invasiveness: Biokinetics, Radiation Dosimetry, and Proof of Concept in Humans

Overexpression of the chemokine-4 receptor (CXCR4) in brain tumors is associated with high cancer cell invasiveness. Recently, we reported the preclinical evaluation of 99mTc-CXCR4-L (cyclo-D-Tyr-D-[NMe]Orn[EDDA-99mTc-6-hydrazinylnicotinyl]-Arg-NaI-Gly) as a SPECT radioligand capable of specifically detecting the CXCR4 protein. This research aimed to estimate the biokinetic behavior and radiation dosimetry of 99mTc-CXCR4-L in healthy subjects, as well as to correlate the radiotracer uptake by brain tumors in patients, with the histological grade of differentiation and CXCR4 expression evaluated by immunohistochemistry. 99mTc-CXCR4-L was obtained from freeze-dried kits prepared under GMP conditions (radiochemical purities >97%). Whole-body scans from six healthy volunteers were acquired at 0.3, 1, 2, 4, 6, and 24 h after 99mTc-CXCR4-L administration (0.37 GBq). Time-activity curves of different source organs were obtained from the image sequence to adjust the biokinetic models. The OLINDA/EXM code was employed to calculate the equivalent and effective radiation doses. Nine patients with evidence of brain tumor injury (6 primaries and 3 recurrent), determined by MRI, underwent cerebral SPECT at 3 h after administration of 99mTc-CXCR4-L (0.74 GBq). Data were expressed as a T/B (tumor uptake/background) ratio. Biopsy examinations included histological grading and anti-CXCR4 immunohistochemistry. Results showed a fast blood activity clearance (T1/2α = 0.81 min and T1/2β = 12.19 min) with renal and hepatobiliary elimination. The average equivalent doses were 6.10E − 04, 1.41E − 04, and 3.13E − 05 mSv/MBq for the intestine, liver, and kidney, respectively. The effective dose was 3.92E − 03 mSv/MBq. SPECT was positive in 7/9 patients diagnosed as grade II oligodendroglioma (two patients), grade IV glioblastoma (two patients), grade IV gliosarcoma (one patient), metastasis, and diffuse astrocytoma with T/B ratios of 1.3, 2.3, 13, 7, 19, 5.5, and 3.9, respectively, all of them with positive immunohistochemistry. A direct relationship between the grade of differentiation and the expression of CXCR4 was found. The two negative SPECT studies showed negative immunohistochemistry with a diagnosis of reactive gliosis. This “proof-of-concept” research warrants further clinical studies to establish the usefulness of 99mTc-CXCR4-L in the diagnosis and prognosis of brain tumors.


Introduction
e aggressiveness of high-grade glioblastomas and the refractoriness to conventional therapies are mostly due to their highly invasive nature [1]. Glioblastomas produce many malignant tissue satellites that frequently migrate to substantial distance from the primary tumors, the reason for which they are challenging to eradicate by surgical techniques and chemo-and radiotherapeutic regimens [2,3].
Previous studies have demonstrated that the overexpression of the chemokine-4 receptor (CXCR4) results in increased migration of glioma tumor cells [4]. is expression is from twenty-five to eighty-nine times higher than that found in noninvasive (low-grade) glioma cells [5]. High expression levels of CXCR4 and its ligand, the chemokine stromal cellderived factor 1-α (SDF1-α � CXCL12), usually indicate a poor prognosis for patients with brain tumors [6]. erefore, CXCR4 and CXCL12 are molecular targets of interest for the development of potential targeted cancer therapies.
is research aimed to estimate the biokinetic behavior and radiation dosimetry of 99m Tc-CXCR4-L in healthy subjects, as well as to correlate the radiotracer uptake by brain tumors in patients, with the histological grade of differentiation and CXCR4 expression evaluated by immunohistochemistry.

Healthy Subjects and
Patients. Six healthy subjects (age range: 28-37 y; mean age ± SD: 33 ± 4 y; 3 females and 3 males) were included. Physical examination and medical history were performed. Individuals with a history of major surgery (e.g., organ removal) or evidence of clinical disease were excluded. Mean (±SD) subject weight was 62 ± 10 kg (range, 51-76 kg). After receiving full information concerning the aims of the study, all volunteers agreed to participate in the complete biokinetic study and signed a consent form. e activity administered to healthy subjects was 0.37 GBq (25 μg of CXCR4-L peptide).
Nine patients between 25 and 72 y (mean age ± SD: 42 ± 16 y), with evidence of brain tumor injury (6 with an indeterminate brain tumor and 3 with suspicion of tumor recurrence), established by magnetic resonance imaging (MRI) studies, were included in the study (Table 1). is research was carried out at the Nuclear Medicine Department at the Specialties Hospital of "Centro Médico Nacional Siglo XXI, Instituto Mexicano del Seguro Social," Mexico. Informed consent was given by the patients, and the protocol was approved by the institutional ethics committee, considering the following aspects: (a) Helsinki Declaration (1975; revised version, 2008) and the ethical standards of the institutional committee related to human experimentation, (b) the GMP certificate granted to ININ by COFEPRIS ("Comisión Federal para la Protección contra Riesgos Sanitarios": regulatory authority in Mexico), (c) the complete preclinical studies of 99m Tc-CXCR4-L, and (d) the basis of "proof-of-concept" and microdosing studies.
2.4. Acquisition of Images. 99m Tc-CXCR4-L images in volunteers were obtained with a Symbia TruePoint dual-head gamma camera (SPECT/CT, Siemens), with high-resolution and low-energy collimators. e established parameters were velocity: 12 cm/min; matrix size: 256 × 1024 pixels; window: 20% symmetric at 140 keV; and scatter corrections: dual-energy window with simultaneous acquisition at 119 keV (20% width). Transmission factors to obtain the body (abdomen and chest) attenuation were calculated by using the I/I 0 counting rate, with (I) and without (I 0 ) the patient of a 99m Tc-filled flood source (555 MBq). Anterior and posterior scintigraphy of the whole body was obtained at 0.3, 1, 2, 4, 6, and 24 h after radiopharmaceutical administration.
In patients, preoperative (15 ± 5 d before surgery) cerebral SPECT (Siemens E. Cam Signature double detector) images were acquired at 3 h after 99m Tc-CXCR4-L (0.74 GBq) administration using a 128 × 128 matrix, window centered on 140 keV, with scattering correction, 360-degree rotation, 128 images of 20 s, and a total duration of approximately 21 min.

Image Analysis.
Visual and semiquantitative analyses were performed by two physicians (specialized in nuclear medicine and molecular imaging) with an experience of >9 years in the use of the Siemens Syngo Acquisition Workplace equipment workstation with processing software for volumetric analysis. e tumor/background (T/Bmax) uptake ratio was calculated by quantifying the number of maximum counts obtained by delimiting volumetric regions (3D) of interest (VOI) with the isocontour around the entire tumor area (T) and in the contralateral brain region (B).

Biokinetic Evaluation.
Images obtained in the DICOM (Digital Imaging and Communication in Medicine) format were processed using ImageJ Software (V1.51i, in Java, Image Processing and Analysis) for scattering correction by using the dual-energy window method. By using the transmission factors (I/I 0 ) experimentally calculated as described above, regions of interest of source organs (liver, bladder, heart, spleen, intestine, lungs, kidneys, and wholebody) were corrected by attenuation. e activity in each source organ was divided by the initial whole-body (WB) activity (100% of injected activity) to determine the injected activity fraction (IA): Technetium-99m time-activity curves were built from the image sequence in each organ. As the heart does not overexpress CXCR4, its activity was considered as having Table 1: Average total number of disintegrations (N) of 99m Tc-CXCR4-L in source organs calculated from six healthy volunteers (3 females and 3 males).   blood activity kinetics. e % IA data at different times were used in the OLINDA/EXM code to calculate the total number of disintegrations (N, MBq . h/MBq). e GI tract model (ICRP 30) included in the code was employed for the excretion model considering an activity fraction of 0.061-0.027 entering the small intestine, as images showed that 4.42 ± 1.69% of the 99m Tc-CXCR4-L injected activity is excreted to the intestine at 20 min after administration (Tables S1-S6, supplementary material). % IA in urine (bladder activity) was also considered as excretion data in the OLINDA code.

99m
Tc-CXCR4-L Absorbed Dose. Equivalent doses were evaluated according to general equation (2), as previously reported [15]: where D target←source is the radiation absorbed, N source is the total number of disintegrations, and DF target←source is the absorbed dose per nuclear transition. e equivalent and effective radiation doses were obtained by using the experimental N values in the OLINDA/EXM code [15].

Tumor Tissue Samples.
All patients underwent total or partial tumor resection. Histopathology was the gold standard to verify the diagnosis as well as to determine the presence of viable tumor tissue. e histopathological reports were collected by the Pathology Department of the "Hospital de Especialidades of CMN Siglo XXI, IMSS," which were interpreted by a certified and experienced   Biosystems). e analysis was performed double-blind by two pathologists, which selected five random fields for each tumor specimen under an optical microscope (magnification 200x). Gastric cancer tissue was used as an internal positive control for the expression of CXCR4.

Statistical Analysis.
To analyze the relationship between the pathological grade (WHO classification) and the brain tumor uptake evaluated by SPECT (T/Bmax ratio), nonparametric tests were used.

Results and Discussion
e radiochemical purity of 99m Tc-CXCR4-L (Figure 1), obtained from the freeze-dried kits, was 98 ± 1%, as obtained by HPLC analyses. e average molar activity was 13 GBq/ μmol before injection to patients.
None of the healthy volunteers reported adverse reactions such as bradycardia, itching, hives, vomiting, nausea, flushing, bronchospasm, dyspnea, coughing, chills, decreased blood pressure, or dizziness after 99m Tc-CXCR4-L administration. e 99m Tc-CXCR4-L blood activity biokinetic model is shown in the following equation: e half-life value was 0.0136 h (0.81 min) for the fast component (T 1/2 α � ln 2/51.01), 0.203 h (12.19 min) for the first slow component (T 1/2 β � ln 2/3.41), and 2.03 h for the second slow component (T 1/2 γ � ln 2/0.34) (Figure 2). e activity was rapidly eliminated by kidneys and the hepatobiliary system (Figure 3). e activity in the kidneys at 20 min was 4.28 ± 1.43%, and after 24 h, it decreased to 0.10 ± 0.07%. Twenty-four hours after the administration of 99m Tc-CXCR4-L, the total activity excreted from the whole body was 99.18 ± 0.24% (Supplementary Material,  Tables S1-S6). e total disintegrations that occurred in each organ are shown in Table 1, and the equivalent doses of 99m Tc-CXCR4-L for the main source organs are shown in Table 2. Due to the fast radiopharmaceutical elimination, the effective dose of 99m Tc-CXCR4-L (2.67-3.14 mSv/740 MBq) is in the same order of magnitude than that reported for other 99m Tc peptides, such as 99m Tc-RGD (6.1 mSv/740 MBq) or 99m Tc-iPSMA (3.42 mSv/740 MBq) [16,17], and below 10 mSv, in agreement with the recommendation of the World Health Organization [18]. Table 3 shows the T/Bmax ratios calculated using SPECT images from patients. Histopathological studies confirmed 3 HGG (1 gliosarcoma and 2 GBM), 2 LGG (oligodendroglioma), 1 metastasis (poorly differentiated carcinoma), 1 recurrent glioma (astrocytoma II), and 2 reactive glioses (see images in the Supplementary Material section). Among the 6 cases with an undetermined brain tumor, SPECT was positive in 3 HGG (T/Bmax 10.8 ± 5.7), 2 LGG (T/Bmax 1.8 ± 0.5), and 1 brain metastasis (T/Bmax 5.5). Of the patients with suspicion of tumor recurrence, SPECT was positive in one (diffuse astrocytoma II) with T/Bmax 3.9 (Figure 4), while for the rest, the result was negative, with a histological report of reactive gliosis. T/Bmax significantly correlated with the pathological grade (WHO classification) (Spearman r s � 0.94, p < 0.003) ( Figure 5).
Immunohistochemical results corroborated that CXCR4 was highly expressed in HGG, whereas its expression in LGG was very low and absent in reactive gliosis ( Figure 6). As it is known, CXCR4 is overexpressed in various tumors, including HGG and brain metastasis, involved in the tumor 10.8 ± 5.7 1.8 ± 0.5 LGG HGG growth, angiogenesis, recurrence, resistance to therapy, shorter survival, and poor conditioning prognosis [6]. In agreement with the results of this research, it has been reported that primary brain tumor cell lines have high concentrations of CXCR4 compared to the normal brain parenchyma, demonstrating CXCR4 overexpression of up to 81% in glioblastoma tumor tissues [19,20]. In this context, Lapa et al. [7]  Tc-CXCR4-L uptake (H), which suggests a diagnosis of gliosis. e biopsy of patient 1 with low-grade glioma (C) shows diffuse cytoplasmic CXCR4 protein immunostaining (+), while the biopsy of patient 4 with high-grade glioma (F) presents intense recognition of CXCR4 in the cytoplasmic and nuclear cell compartments (+++). Gliosis tissue of patient 9 (I) showed no HRP (horseradish peroxidase) staining, indicating the absence of CXCR4 cell expression, which correlates with the negative 99m Tc-CXCR4-L uptake. uptake, which allowed adequate discrimination between tumor recurrence and gliosis. Likewise, the absence of intraparenchymal concentration permitted to obtain images for clinical use with appropriate T/Bmax ratios (HGG 10.8 ± 5.7, LGG 1.8 ± 0.5, and cerebral metastasis 5.5). e immunohistochemical confirmation of higher CXCR4 expression in patients with HGG with regard to those with LGG corroborated the directly proportional relationship between the grade of differentiation and the expression of CXCR4. is "proof-of-concept" research showed the potential utility of 99m Tc-CXCR4-L in the specific diagnosis and prognosis of brain tumors.

Conclusions
e equivalent and effective radiation doses of 99m Tc-CXCR4-L in humans are comparable to those from most diagnostic studies of 99m Tc. is preliminary study warrants further clinical studies in order to establish the usefulness of 99m Tc-CXCR4-L in the diagnosis, prognosis, prediction of therapeutic response, discrimination between recurrence, and reactive gliosis, as well as in the selection of patients who could benefit from targeted anti-CXCR4 therapies.

Data Availability
e data used to support the findings of this study are included within the article.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.