¹⁸F-Fluorodeoxyglucose positron emission tomography may not visualize radiation pneumonitis

To confirm the feasibility of the proposed experiments, 14 male Wistar rats underwent CT-based treatment simulation, treatment planning, and treatment delivery, followed by weekly CT scanning and pathologic evaluation of lung tissues to confirm the presence of pneumonitis. RT simulation was done while the rats were restrained in the prone position in a specialized device, with 3 markers used to determine the radiation target (Additional file 1: Fig. S1). CT images were obtained with a Philips Brilliance Big Bore system with bellows (Philips Healthcare, DA Best, The Netherlands) and transmitted to a Varian Eclipse 10.0 Treatment Planning System. Every rat had an independent RT plan, for which a radiation oncologist contoured the entire right lung, and a physicist developed the plan. RT was delivered with 6-MV X-rays, to a total dose of 20 Gy in a single fraction, with a Varian Unique Medical Linear Accelerator while the rats were anesthetized with 2% pentobarbital sodium at a dose of 30 mg/kg (Sigma-Aldrich Co. LLC, St. Louis, MO, USA). In fact, this dose was selected according to the previous published studies, in which they declared that a single dose of 20 Gy develops considerable radiation lung injury [17, 18]. The left lung was shielded to minimize radiation exposure. Target regions were irradiated with a margin of at least 2 mm from the center to prevent radiation esophagitis or myelitis. Other regions were protected from irradiation with a multileaf collimator. CT scans were obtained weekly to monitor the progress of radiation pneumonitis, and at 1 day, 1 week, 2 weeks, 4 weeks, 6 weeks, and 7 weeks after the RT, two rats per treatment group (see below) were killed and their lungs examined for pathological evidence of radiation-induced lung injury. We determined the 7-week study period through weekly CT scanning and pathologic evaluation. Under our experimental conditions, almost all the rats could develop radiation pneumonitis at this time point, which was also consistent with the 6–8 weeks study period reported in the previously published literature [17‐20]. Radiation pneumonitis refers to interstitial pneumonitis that occurs after radiotherapy, not caused by other factors such as bacterial infection, usually manifested as diffuse alveolar wall thickening and the accumulation of inflammatory cells.

In the formal experiment, 40 male Wistar rats weighing 178–220 g each were randomly assigned to one of four treatment groups (10 rats/group): control, RT-only, LPS-only, and RT plus LPS. RT was planned and delivered as described above. LPS (obtained from Solarbio Life Sciences, Beijing, China) was given as a single intranasal injection of 5 mg/kg body weight [10] at 24 h before the micro-PET scan. All rats had micro-PET scans at 7 weeks after RT (or sham). All experiments were repeated twice, and all findings were similar across all experiments.

¹⁸F-FDG was synthesized according to the method developed by Ido [21] to a purity of > 95% in the Jiangsu Atomic Energy Laboratory (Jiangsu, China). Micro-PET images were acquired with an Inveon PET scanner (Siemens Preclinical Solutions, LLC, Knoxville, TN, USA). Before scanning, rats were fasted for 8–12 h, after which 7.4–11.1 MBq of ¹⁸F-FDG was injected via the tail vein and PET scans were obtained 60 min later. Anesthesia was maintained during the PET scans with 1.5% isoflurane in 100% oxygen at 1.5 L/min. For the scans, rats were placed prone on the bed of micro-PET, and the limbs were fixed with tape. A 10-min static single-frame scan was acquired with a small-animal PET camera, and images were reconstructed by ordered subsets expectation maximization (OSEM)-3D IAW (Siemens Preclinical Solutions).

Two experienced nuclear medicine physicians examined all PET images in a double-blinded fashion. Regions of interest (ROIs) in the right lungs were drawn using vendor software (IS_v1.4.3 SP1; Siemens Healthineers, Erlangen, Germany). The SUV was calculated as an absolute measure of ¹⁸F-FDG uptake in an ROI as:

[(measured activity concentration, in kBq/mL)/(injected dose, in kBq/body weight, in g)].

All animals were weighed before scanning. SUV_max and SUV_mean were defined as the maximum and mean tracer uptake in the ROIs.

After the micro-PET scan, rats were killed by anesthetizing, and the right lungs were removed, fixed overnight in 10% formalin, and embedded in paraffin. Serial 4-μm tissue sections were prepared for histologic analysis and immunohistochemical staining of PKM2 and GLUT1, detailed information was provided in the Additional file 1. Digital images of the immunohistochemical-stained slides were acquired with a Nikon Eclipse 80i microscope equipped with Nikon DS-Fi1 camera. All images were taken under high magnification (× 400) with consistent imaging parameters (uniform light source, exposure time, and autofocus). Expression levels of PKM2 and GLUT1 were measured with Image-Pro Plus (version 6.0, Media Cybernetics, Rockville, MD, USA) and quantified as integrated optical density (IOD)/area.

Concentrations of interleukin (IL)-1, IL-6, and transforming growth factor (TGF)-β in serum and bronchoalveolar lavage fluid (BALF) from the right (irradiated) and left (unirradiated) lungs were measured with a Rat IL-1 ELISA Kit, a Rat IL-6 ELISA Kit, and a Rat TGF-β ELISA Kit (Abcam, Cambridge, MA, USA) according to the manufacturer’s instructions. Plates were read with a VersaMaxMultiplate spectrophotometric reader at 450 nm.

Lactate content (mmol/L) = [A₅₃₀(sample) – A₅₃₀(blank)]/[A₅₃₀(standard) – A₅₃₀(blank)] × standard concentration (3 mmol/L) × sample dilution multiple;

Pyruvate content (μmol/mL) = [A₅₀₅(sample) – A₅₀₅(blank)]/[A₅₀₅(standard) – A₅₀₅(blank)] × standard concentration (0.2 μmol/mL) × sample dilution multiple.

Statistical analyses were done with SPSS 23.0 (SPSS Inc., Chicago, IL, USA). Differences between two groups were assessed with two-tailed paired t tests. One-way analysis of variance and Bonferroni’s post hoc test were used for multiple comparisons. Pearson correlation analysis was used to evaluate associations between groups. The threshold for statistical significance was set at P < 0.05.

In the pre-experiment to determine the rat model of radiation pneumonitis, significant radiation pneumonitis was apparent on weekly micro-CT scans by 7 weeks after RT (Additional file 1: Fig. S2). Pathologic changes during this period are shown in Additional file 1: Fig. S3. At 7 weeks after RT, plasma levels of IL-1α and TGF-β in irradiated rats were significantly higher than in non-irradiated rats (P = 0.006 and P = 0.006), but receipt of RT did not affect plasma IL-6 levels (P = 0.951) (Additional file 1: Fig. S4). In irradiated rats, BALF samples from the right (irradiated) lung had higher levels of IL-1α, IL-6, and TGF-β than in BALF from the left (non-irradiated) lung (P = 0.014, P = 0.027, and P = 0.008, respectively). However, in non-irradiated rats, no differences were found in IL-1α, IL-6, or TGF-β in BALF from the left and right lungs (P = 0.819, P = 0.624, and P = 0.832, respectively) (Fig. 1).

Micro-PET scans were obtained from all 40 rats (10 per group). No significant FDG uptake was observed in the RT-only and control groups, but the RT+LPS and LPS-only groups had obvious FDG uptake in the right lung (Fig. 2). The SUV_max value was highest in the RT+LPS group (11.87 ± 1.16) and was significantly higher than that in the LPS group (8.16 ± 1.53, P < 0.001), the RT-only group (4.53 ± 0.87, P < 0.001), and the control group (3.99 ± 0.75, P < 0.001). SUV_max in the LPS-only group was also higher than that in the RT-only and control groups (P < 0.001 for both), but was not different in the RT-only and control groups (P = 0.156) (Fig. 3a). Similar results were also found for SUV_mean, with values of 2.87 ± 0.55 (control), 3.10 ± 0.44 (RT-only), 5.57 ± 0.70 (LPS-only), and 8.30 ± 0.75 (RT+LPS). Compared with the control group, SUV_mean was significantly elevated in the RT+LPS and LPS-only groups (P < 0.001 for both), but were not different in the RT-only group (P = 0.304) (Fig. 3b).

To explore mechanisms that could affect FDG uptake, we investigated two important proteins in the aerobic glycolysis pathway, PKM2 and GLUT1 (Fig. 4). PKM2 is the main rate-limiting enzyme in glycolysis, and its methylation acts to switch from oxidative phosphorylation to aerobic glycolysis. In the current study, PKM2 was increased in the LPS-only and RT+LPS groups compared with control (P = 0.018 and P < 0.001), but RT-only did not affect PKM2 expression (P = 0.270) (Fig. 5a). GLUT1 is a transporter on the surface of cell membranes that can transport glucose directly into cells. Compared with control, GLUT1 expression was higher in the RT+LPS group (P = 0.004) but was not different in the LPS-only (P = 0.068) or RT-only (P = 0.989) groups (Fig. 5b).

We also assessed levels of lactate, the end product of aerobic glycolysis, in serum samples. Mean serum lactate levels were 5.36 ± 0.21 mmol/L in the control, 5.93 ± 0.57 mmol/L in the RT-only, 6.45 ± 0.38 mmol/L in the LPS-only, and 7.85 ± 0.66 mmol/L in the RT+LPS groups; serum lactate levels were higher in the RT+LPS group than in the control (P = 0.002) or RT-only (P = 0.021) groups but were not different from those in the LPS-only group (P = 0.084) (Fig. 6a). We also measured serum levels of pyruvate, an intermediate product of glucose metabolism that can be used to complete the tricarboxylic acid cycle or aerobic glycolysis, and found that serum pyruvate levels did not change after RT-only or LPS-only (P = 0.617 or P = 0.562) but were higher in the RT+LPS group (P = 0.024 vs. control) (Fig. 6b).

Pearson correlation analysis indicated that SUV_max was strongly related to PKM2, GLUT1, and lactate (r = 0.71 P < 0.001 [PKM2], r = 0.64 P < 0.001 [GLUT1], and r = 0.52 P < 0.001 [lactate]) (Fig. 7). Similar results were found for SUV_mean (r = 0.71 P < 0.001 [PKM2], r = 0.58 P < 0.001 [GLUT1], and r = 0.51 P < 0.001 [lactate]). PKM2, the main rate-limiting enzyme in glycolysis, correlated with GLUT1 (r = 0.52, P < 0.001) and lactate (r = 0.53, P < 0.001). Pyruvate, an intermediate in aerobic glycolysis, was not correlated with other factors, especially GLUT1 (r = − 0.073, P = 0.814). In sum, both ¹⁸F-FDG uptake measurements (SUV_max and SUV_mean) correlated strongly with immunohistochemical markers and lactate levels, suggesting that the Warburg effect had been triggered in response to LPS, but not radiation.