Breast cancer (BC) is a heterogeneous disease with several subtypes including triple-negative breast cancer (TNBC), which has a poor prognosis.1 This classification is defined by the absence of estrogen and progesterone receptor expression, and the nonoverexpression or amplification of human epidermal growth factor receptor 2 (HER2). Recently, the combination of immune checkpoint inhibitors targeting programmed death 1 (PD-1) or programmed death ligand 1 (PD-L1) with chemotherapy has improved pathological complete response rates (neoadjuvant treatment) and long-term outcomes in nonmetastatic (neoadjuvant and adjuvant treatment) and metastatic TNBC.2–5

Despite these advances, a subset of TNBC patients does not exhibit a favorable response to these therapeutic regimens. Consequently, there is a compelling need for strategies that can predict TNBC status to improve the outcome of immune-targeted therapies.

Immunohistochemical (IHC) approaches have shown that the TNBC tumor microenvironment (TME) is characterized by high expression of growth and migration-promoting molecules, and by a greater number of immune cells, especially by tumor-associated macrophages (TAMs).6 In fact, several studies have shown that the TME plays a key role in tumor development and response to treatment,7 and TAMs may play a key role in tumor progression and treatment-resistant mechanisms.8 Indeed, TAMs have been shown to modulate PD-1/PD-L1 expression and increase PD-L1 expression via cytokine secretion.9 It is therefore essential to study TME in TNBC and its interaction with tumor cells, but currently, we only have IHC approaches available, which require biopsy analysis, and it would be beneficial to have a comparable noninvasive characterization method.

Molecular imaging could play a role in the noninvasive study of TNBC TME. Indeed, the translocator protein 18 kDa (TSPO) is expressed by TAMs and can be targeted by PET using the radiopharmaceutical 18F-DPA-714 (N,N-diethyl-2-[4-(2-fluoroethoxy)phenyl]-5,7-dimethylpyrazolo[1,5-a]pyrimidine-3-acetamide labeled with fluorine 18). TSPO, originally described as the peripheral benzodiazepine receptor, is an intracellular protein involved in many biological functions including apoptosis, cell proliferation, and immunomodulation.10 TSPO is predominantly localized on the outer mitochondrial membrane, but it is also found around the nuclei of certain tumor cells, particularly in BC.11 Although this ubiquitous protein plays essential physiological roles, it is also involved in various pathological phenomena, including tumor progression. Increased TSPO protein expression has been observed in numerous human tumor cells and tissues, particularly in hormone receptor-negative BCs, where TSPO expression appears to correlate with tumor aggressiveness and progression.12–14 Additionally, TSPO serves as an in vivo biomarker of peripheral inflammation and is particularly expressed in myeloid lineage cells, such as macrophages.15,16 TSPO PET/CT has previously been used as a biomarker of brain neuroinflammation,17 for assessing joint inflammation in rheumatoid arthritis18 or for evaluating lung inflammation.19

Overall, TSPO is emerging as a promising in vivo prognostic and therapeutic target. A preclinical study using human breast tumor cell transplantation models in mice and 18F-DPA-714-based imaging has validated TSPO as a biomarker of tumor cells and TME, particularly for TAMs.20

The aim of this pilot feasibility study (EITHICS) was to compare 18F-DPA-714 PET/CT imaging with tumor data obtained by IHC, autoradiography, and TSPO polymorphism to investigate the primary operable TNBC tumor and its TME.

PATIENTS AND METHODS Design and Population

This study (EITHICS) is a prospective, multicenter, nonrandomized phase II trial. Patients were enrolled between June 2020 and May 2021 in 2 hospitals in France. Ethical approval was obtained from the Ile de France II Review Board (19.10.17.46746). All patients signed a written informed consent prior to participation in this study (2019-10-00030).

All patients had TNBC (estrogen and progesterone receptors both <10%, and HER2 not amplified or overexpressed) with a surgical indication. Patients with inflammatory or metastatic TNBC were excluded. Patients were also excluded if they had received any anticancer treatment (chemotherapy, immunotherapy, hormone therapy, and external beam radiotherapy), antibiotics, and/or anti-inflammatory drugs (steroidal and/or nonsteroidal) in the 30 days prior to inclusion.

The study design is shown in Supplemental Figure 1, https://links.lww.com/CNM/A476. Initially, a blood sample was taken from each patient for TSPO gene polymorphism detection (rs6971). Because of the time required to obtain the genotyping results, a pooled and staggered analysis was chosen at the end of the study. The TSPO polymorphism (rs6971) predicts the binding affinity to second-generation TSPO tracers,21 including 18F-DPA-714, and patients included were potentially high, medium, or low affinity binders (HAB, MAB, and LAB). The TSPO gene polymorphism status was determined for all patients.

Whole-body 18F-FDG PET/CT and multiparametric MRI with bilateral breast and axillary diffusion sequence were performed initially. For the 18F-DPA-714 PET/CT scan, 3 sequential acquisitions were performed. After surgery, tumor tissue was examined by IHC analysis, and in vitro autoradiography using 3H-DPA-714 and 3H-PK-11196 ((R)-N-methyl-N-(1-methylpropyl)-1-(2-chlorophenyl)isoquinoline-3-carboxamide).

Radiosynthesis of 18F-DPA-714

DPA-714 was labeled with fluorine-18 at its 2-fluoroethyl moiety, following nucleophilic substitution of the corresponding tosylate analog, according to slight modifications of previously reported procedures.22 The formulation of 18F-DPA-714 provided a sterile injectable solution of isotonic sodium chloride with ethanol at a mass percentage of <8% for total injected volumes (ranging from 3 to 5 mL), in accordance with the European Pharmacopeia. The mean specific activity of 18F-DPA-714 obtained was 251.65 ± 60.23 GBq/μmol.

Imaging Acquisition Parameters 18F-DPA-714 Images

The radiopharmaceutical activity delivered was 3.5 MBq/kg ± 10%, in agreement with the literature.17 PET images were acquired with 2 types of cameras: Biograph Vision 450 and Biograph mCT (Siemens Healthcare, Erlangen, Germany). A first series of dynamic PET images in the procubitus position was acquired in list mode for 30 minutes (30 images of 1 minute each), centered on the thorax and axillary area, starting at the time of injection. The PET acquisition was combined with a low-dose CT scan centered on the chest. Then, a 15-minute thoracic static acquisition in procubitus was performed at 45 minutes postinjection. Finally, a decubitus whole-body acquisition from the head to the mid-thighs was performed 60 minutes postinjection, 120 kV, 60 mAs, with 5-mm slice thickness (16 × 1.2 mm), and a pitch of 0.85.

MRIs

Procubitus breast MRI was performed with a dedicated 16-channel breast coil on a 1.5 T MAGNETOM SOLA (Siemens Healthcare GmbH, Erlangen, Germany). The protocol included preinjection axial T1 TSE and T2 TSE, 3 mm. After automatic injection of Clariscan GE 0.5 mmol/mL (gadoteric acid), 0.2 mL/kg, followed by a 20-mL saline bolus, a DIXON VIBE 3D iso 0.9 dynamic axial acquisition was performed, followed by 4 postinjection series.

Data Analysis

For each patient, on 18F-DPA-714 PET/CT fusion images (procubitus dynamic images sum), a volume of interest (VOI) on the breast tumor lesion was defined using a thresholding method relative to 40% tumor SUVmax, using 3D Slicer software (version 5.2), then transposed onto the healthy contralateral mammary gland and contralateral pectoral muscle to obtain time-activity curves (TACs). These TACs allowed the analysis of 18F-DPA-714 kinetics over the first 30 minutes of PET acquisition for the different tissues. In addition, a VOI was manually traced around the tumor, including the peritumoral tissue. The ratios (tumor-to-background ratio or TBR) between the mean activity SUVmax of the breast tumor lesion VOI and the healthy breast tissue VOI were calculated. The SUVmax values of the tumor and contralateral breast tissue were determined manually by 2 experienced nuclear medicine physicians using the syngo.via 3D postprocessing software (Siemens Healthcare GmbH, Erlangen, Germany).

Histologic and IHC Evaluations

Thirteen cases of archival formalin-fixed and paraffin-embedded tissues of the internal cohort were evaluated on full sections for histological evaluation. All tumor tissue samples were surgically collected and were fixed in 10% neutral buffered formalin for standard histological analysis and IHC. IHC was performed using the Leica BOND-MAX platform. Details of the antigen retrieval technique and dilution of primary antibodies CD68 and CD163 are listed in Supplemental Table 1, https://links.lww.com/CNM/A477. The number of CD68-positive and CD163-positive cells per mm2 was evaluated in hot spots. Tumor-infiltrating lymphocytes and CD138-positive plasma cells were assessed according to recommendations of an international working group.23,24

Comparative Evaluation of TSPO Radioligands 3H-DPA-714 and 3H-PK-11195 by In Vitro Autoradiographic Binding on TN Tumor Slices

The density-specific TSPO binding sites were measured by in vitro autoradiographic assessments using 3H-DPA-714 and 3H-PK-11195 on adjacent TNBC tumor sections from 10 of the 13 included patients (the surgical specimen obtained from tumor resection surgery was insufficient for the 3 other patients). 3H-DPA-714 was prepared according to Damont et al,25 (molar activity 2.1 GBq/μmol, CEA, Orsay, France), and 3H-PK-11195 was commercially purchased (molar activity 3.06 GBq/μmol; Perkin Elmer, Norwalk, CT). Frozen samples were sliced into 16-μm-thick sections using a cryostat at −20°C (CM 3050S; Leica, Germany). Sections on gelatinized slides were stored at −80°C until required. Tumor sections were allowed to equilibrate at room temperature (RT) for 3 hours, then incubated with 1 nM of labeled ligand (3H-DPA-714 or 3H-PK-11195) in 50 mmol/L Tris–HCl buffer pH 7.4 at RT for 60 minutes. Nonspecific binding was assessed by incubation of adjacent sections in the presence of 1 μM of cold PK-11195 (Sigma Aldrich, Lyon, France). Sections were rinsed twice in ice cold buffer (4°C) for 5 minutes, then briefly in distilled water at 4°C and dried at RT. Slides were made conductive by applying a metal electric tape (3M; Euromedex, Souffelweyersheim, France) and then were placed in the gas chamber of the b-Imager 2000 (Biospace Lab, Paris, France). Acquisitions were compiled over a period of 3 hours. Hyperintense binding areas were manually contoured on each total binding section, and each defined region of interest was reported on adjacent slices pretreated with cold PK11195 for nonspecific binding measurement. The level of bound radioactivity was quantitated by counting the number of β-particles emitted from the delineated area (β-Vision Software; Biospace Lab, Paris, France) in the regions of interest. The radioligand signals, expressed in counts per minute per square millimeter (cpm/mm2), were measured for 6 sections per patient using an image analyzer (M3 Vision; Biospace Instruments, Paris, France). Specific binding (SB) was determined by subtracting nonspecific binding from total binding. The 3H-DPA-714 SB/3H-PK-11195 SB ratio was determined for each patient to evaluate the signal intensity for each radioligand, relative to TSPO genotype.

Statistical Analysis

Qualitative factors were described by the frequency of their respective modalities and continuous factors by their mean ± standard deviation (or median, range). Post hoc subgroups were compared using Pearson χ2 test (or Fisher test if necessary) for categorical variables and Wilcoxon test (or Kruskal-Wallis or Mann-Whitney U tests if necessary) for continuous variables. Correlations between continuous variables were calculated using Spearman correlation coefficient. All tests were 2-tailed, with significance set at 5%. The software used was Stata18.0 (StataCorp, College Station, TX).

RESULTS

A total of 13 TNBC patients were included in the study. Breast MRI confirmed unifocal breast involvement in all patients. The median tumor size was 18 mm (range, 9–33) with a median Ki67 of 60% (range, 13%–90%). Patient characteristics are summarized in Table 1. Eleven patients underwent lumpectomy, and 2 underwent mastectomy. Ten of 13 of patients had tumors classified as pT1, and 3/13 were classified as pT2. Lymph node involvement was histologically identified in only 1 patient.

TABLE 1 - Patient Characteristics # Age Histological Type cTNM Surgery Type SALND Lesion Size (mm) pTpN SBR Grade UICC Status % ER Status % PR Status HER2 Status Vascular Emboli Ki67 (%) 1 51 NSIC T1N0M0 Mastectomy Yes 16 T1N0 III I 0 0 1+ No 75 2 68 NSIC T1N0M0 Lumpectomy Yes 20 T1N1 II IIA 0 0 0 No 22 3 69 NSIC T1N0M0 Lumpectomy Yes 18 T1N0 III I 0 0 0 Yes 35 4 75 NSIC T2N0M0 Lumpectomy Yes 23 T2N0 II IIA 5 5 0 Yes 13 5 41 NSIC T2N1M0 Lumpectomy Yes 13 T1N0 II I 0 0 0 No 20 6 49 NSIC T1N0M0 Lumpectomy Yes 26 T2N0 III IIA 0 0 0 No 80 7 68 NSIC T0N0M0 Lumpectomy Yes 16 T1N0 III I 0 0 1+ Yes 60 8 54 NSIC T1N0M0 Lumpectomy Yes 10 T1N0 III I 0 0 0 No 70 9 50 NSIC T1N0M0 Lumpectomy Yes 33 T2N0 III IIA 0 0 2+ (Fish neg) No 90 10 67 NSIC T1N0M0 Lumpectomy Yes 18 T1N0 III I 0 0 0 No 40 11 26 NSIC T1N0M0 Lumpectomy Yes 18 T1N0 - I 0 0 0 No 60 12 84 IPC T1N0M0 Mastectomy Yes 20 T1N0 III I 0 0 0 Yes 40 13 53 NSIC T1N0M0 Lumpectomy Yes 9 T1N0 III I 0 0 2+ (Fish neg) No 60

cTNM, clinical stage TNM; SBR grade, Scarff-Bloom-Richardson grade; UICC, Union for International Cancer Control; SALND, sentinel axillary lymph node dissection; pTpN, pathological stage TNM; % ER, % estrogen receptor; % PR, % progesterone receptor; NSIC, nonspecific invasive carcinoma; IPC, invasive papillary carcinoma.

Histological data showed that almost all patients had nonspecific infiltrating carcinoma (12/13, 92%). No adverse events occurred immediately or remotely after 18F-DPA-714 injection. No relapses or deaths were observed during a median follow-up of 33.4 months (range, 27.6–38.5).

18F-DPA-714 PET/CT showed significant tumor uptake with a good TBR, except in 2 patients. Median lesion SUVmax values were 4.73 (range, 0.98–11.09) for 18F-DPA-714 PET/CT and 7.17 (range, 2.27–33.84) for 18F-FDG PET/CT. All lesion SUVmax and SUVmean values for 18F-DPA-714 and 18F-FDG are detailed for each patient in Supplemental Table 2, https://links.lww.com/CNM/A478. The imaging panel performed for each patient, MRI, 18F-DPA-714 PET/CT, and 18F-FDG PET/CT, is shown in Figure 1. In all patients, 18F-DPA-714 PET/CT showed physiological uptake of bone marrow limited to the spine, spleen, myocardium, salivary glands, gallbladder, liver, and, to a lesser extent, digestive structures and muscles. Apart from these 18F-DPA-714 foci and the tumor involvement, there was no evidence of other uptake, in particular in the brain or in the joints.

FIGURE 1:

Patient 4 iconography: mpMRI (1) dynamic early 18F-DPA-714 PET/CT acquisition on procubitus (2A), whole-body 18F-DPA-714 PET/CT on decubitus at 60 minutes postinjection (2B), and whole-body 18F-FDG PET/CT (3).

Figure 2 shows 18F-DPA-714 TACs uptake for all subjects in the tumor, breast background (contralateral healthy breast) and muscular background (tumor contralateral pectoral). Three different patterns of TACs of 18F-DPA-714 binding in TNBC were observed and classified as “above,” “equal to,” and “below” muscular background. Healthy breast TAC was almost flat (ie, baseline levels), in contrast to pectoral muscle TAC, which showed an 18F-DPA-714 rapid uptake with a slope of 0.55 followed by an asymptotic appearance with a median SUVmean of 1.18 (range, 0.18–3.88). Three different tumor TAC patterns were observed, and we chose to compare them with the muscular background. From the TACs obtained for the “above muscular” group (2 HAB and 2 MAB), the “equal muscular” group (3 HAB, 3 MAB, and 1 LAB), and the “below muscular” group (1 LAB and 1 MAB), the slope was calculated, and the values found were 1.35, 0.62, and 0.22, respectively. For the “above muscular,” “equal muscular,” and “below muscular” groups, the plateau of each curve had a median SUVmean of 4.02 (range, 2.09–5.31), 1.66 (range, 0.93–3.07), and 0.61 (range, 0.43–1.02), respectively.

FIGURE 2:

Time-activity curves of 18F-DPA-714 binding on breast background (contralateral healthy mammary gland) and muscular background (pectoral). Three different patterns of TACs of 18F-DPA-714 binding on TNBC were observed: “above,” “equal to,” and “below” muscular background.

For the 13 patients considered together, the calculated TBR (SUVmax TNBC/SUVmax contralateral breast) at the 15th min postinjection was 11.71 (range, 2.99–51.15). For each “above,” “equal,” and “below” group, the TBRs were 19.94 (range, 17.61–51.15), 10.70 (range, 7.32–17.96), and 3.77 (range, 2.99–4.54), respectively. These data are supported by the TACs, which showed a significant difference between the breast background and the TNBC curves.

Peritumoral macrophage infiltration was assessed by IHC using the CD68 marker (Fig. 3). The percentage of M2-polarized macrophages was estimated using the CD163 marker (Table 2). The majority of TNBC tumors (11/13, 84%) showed a preponderance of M2-polarized macrophages with a median of 82% (range, 44%–94%). No significant correlation was observed between the semiquantitative TNBC lesion parameters (SUVmax, SUVmean, and %ID/g) of 18F-DPA-714 PET/CT and the number of CD68 macrophages per field estimated by IHC (P = 0.311) or the percentage of M2-polarized macrophages estimated by IHC using the CD163 marker (P = 0.122).

FIGURE 3:

Immunohistochemical staining of 2 TNBC cases. Anti-CD68 staining of breast tumor biopsy showing high (A) and low (B) CD68 macrophage abundance. Magnification ×100.

TABLE 2 - Intratumoral Immunological Quantification of CD163 and CD68-Positive Macrophages # Nb CD68 Positive-Cells per mm2 (M1 and M2 Type Macrophages) Nb CD163 Positive-Cells per mm2 (%) (M2 Type Macrophages) 1 88 83 (94) 2 27 17 (63) 3 62 51 (82) 4 9 4 (44) 5 47 39 (83) 6 77 63 (82) 7 60 52 (87) 8 12 6 (50) 9 85 74 (87) 10 114 102 (89) 11 43 32 (74) 12 24 15 (62) 13 37 27 (73)

The results for the TSPO polymorphism screening are shown in Table 3 and include the comparative evaluation of TSPO radioligands 3H-DPA-714 and 3H-PK-11195 by in vitro autoradiographic binding on TNBC tumor slices. It is noteworthy that PK11195 is not sensitive to TSPO polymorphism, unlike second-generation TSPO radioligands, including DPA-714.21 First, the homozygous rs6971 polymorphism (A/A) was found in 2 of 13 (15%) patients, which is related to a “low binding” status (LAB) for the second-generation TSPO radiopharmaceuticals as 18F-DPA-714. The other patients were either heterogeneous (MAB) (A/G; n = 6; 46%) or lacked polymorphism (HAB) (G/G; n = 5; 39%). Binding of 3H-PK-11195 and 3H-DPA-714 to TSPO was compared using autoradiography on TNBC tumor sections from 10 patients (2 HAB, 6 MAB, and 2 LAB). Autoradiography quantitative results revealed that SB of the radioligand was significantly higher with 3H-PK-11195 than with 3H-DPA-714 (P < 0.01). In addition, no statistically significant correlation was found with DPA-714 between in vivo PET/CT SUVmax and SUVmean and in vitro SB intensity. Interestingly, and despite the small sample size, the 3H-DPA-714 SB/3H-PK-11195 SB ratio was significantly lower when comparing LAB patients with HAB and MAB patients pooled together (P < 0.05).

TABLE 3 - TSPO Polymorphism and Comparative Evaluation of TSPO Radioligands 3H-DPA-714 and 3H-PK-11195 by In Vitro Autoradiographic Binding on TNBC Tumor Slices # TSPO Polymorphism (1) 3H-PK-11195 SB (cpm/mm2) (2) 3H-DPA-714 SB (cpm/mm2) Ratio (2)/(1) 1 High affinity binder (HAB) – – – 2 Intermediate affinity binder (MAB) 6.16 2.86 0.46 3 High affinity binder (HAB) – – – 4 Intermediate affinity binder (MAB) 5.20 3.31 0.64 5 Intermediate affinity binder (MAB) 9.10 6.97 0.77 6 High affinity binder (HAB) 7.85 5.83 0.74 7 Low affinity binder (LAB) 8.70 2.71 0.31 8 High affinity binder (HAB) – – – 9 High affinity binder (HAB) 7.32 4.3 0.59 10 Low affinity binder (LAB) 10.60 3.69 0.35 11 Intermediate affinity binder (MAB) 8.75 4.72 0.54 12 Intermediate affinity binder (MAB) 15.04 6.79 0.45 13 Intermediate affinity binder (MAB) 13.01 9.61 0.74

Qualitative autoradiographic analysis showed that uptake in some tumor sections was more intense at the tumor periphery than in the center, particularly with 3H-PK-11195, as if uptake was stronger in the TME than in the tumor cells themselves.

DISCUSSION

To our knowledge, this proof-of-concept clinical study is the first to assess the feasibility and relevance of evaluating TNBC primary tumors at the time of diagnosis using PET/CT with 18F-DPA-714, a second-generation TSPO tracer.

Two important findings should be highlighted: first, 18F-DPA-714 PET/CT clearly identified TNBC tumors with an excellent TBR, and the TACs generated for each patient showed that 18F-DPA-714 uptake distinguished 3 significantly different patterns. Second, the results of the TSPO polymorphism and the comparative evaluation of the TSPO radioligands 3H-DPA-714 and 3H-PK-11195 by in vitro autoradiography binding to TNBC tumor sections showed that the ratio (3H-DPA-714 SB)/(3H-PK-11195 SB) was significantly lower when comparing LAB patients with HAB and MAB patients pooled together, which, despite the small sample size, seems to be consistent with the PET/CT data.

With regard to the 18F-DPA-714 biodistribution, similar to Arlicot et al,17 we observed intense physiological uptake in the myocardium, spleen, bone marrow (spine), liver, kidneys, salivary glands, adrenals, and, to a lesser extent, muscular structures. We also observed intense binding in the gallbladder and digestive structures, which is compatible with partial hepatobiliary elimination of the radiotracer, as suggested by the same team.17 Despite high physiological uptake, there may be a gradient of uptake in favor of the neoplastic lesion, making it possible to distinguish between physiology and neoplasia. We were unable to document this possible aspect as all our patients were imaged in the early stages of their disease. However, all these physiological foci could limit the usefulness of 18F-DPA-714 PET/CT for theranostic purposes.

Qualitative analysis of 18F-DPA-714 PET/CT showed significant uptake in breast tumor lesions with an excellent TBR, except in 2 patients. One of these patients had a TSPO LAB polymorphism, described in the literature outside oncology, particularly in neurodegenerative diseases, as corresponding to 10% of the population. The TSPO LAB polymorphism appeared to be nonbinding to the second-generation TSPO tracer such as 18F-DPA-714.17,26 Indeed, our autoradiographic results suggested reduced binding of 18F-DPA-714 to TSPO in TNBC patients with an LAB polymorphism compared with MAB + HAB patients, based on the ratio (3H-DPA-714 SB)/(3H-PK-11195 SB), which was significantly lower in LAB patients. Thus, this observation suggests a different pattern of in vivo TSPO binding between LAB and MAB + HAB patients. Although these data are consistent with the results of brain studies, the literature also reports that 18F-DPA-714 PET/CT can visualize arthritic joints independently of polymorphism status, suggesting a lesser impact of TSPO polymorphism status on arthritis targeting.18 The other patient, with low tumor contrast, had a TSPO MAB polymorphism but the smallest tumor in the sample, measuring only 9 mm on histological analysis, which may partly explain the lack of significant lesion uptake due to the partial volume effects.

TSPO expression in BC cell lines has been extensively studied by Zheng et al using DPA-714 autoradiography. The authors did not report any heterogeneity in radioactivity distribution between the tumor itself and its microenvironment but noted a strong correlation with F4/80-positive macrophage cell expression, an IHC biomarker of an inflammatory microenvironment.20 More recently, Tuominen et al14 reported the same pattern of DPA-714 binding in a head and neck cancer xenograft autoradiography study. In both studies, DPA-714 radiolabeled with fluorine-18 was used for autoradiographic purposes. Of interest, in the present study, we used tritiated DPA-714, which has a lower beta energy and is therefore more suitable for autoradiographic experiments, with improved resolution and quantification of the signal. This overexpression of TSPO in both the neoplastic and stromal compartments of the tumor could therefore contribute significantly to the sensitivity of the PET signal.

The 18F-DPA-714 TACs obtained for TNBC patients showed 3 different kinetic patterns compared with the kinetics observed for the contralateral pectoral muscle. Ribeiro et al27 has already shown, in a completely different pathology (cerebral st