2.1 Animals

A total of 36 adult male (8 to 15 months) Tibet miniature pigs (henceforth mini-pigs) were used for TBI exposure. The pigs were maintained in the animal maintenance facilities and all animal studies were performed under the guidelines and protocols of the Institutional Animal Care and Use Committee of authors’ institution (Permit Number: 2010040). All surgery was performed under ketamine anesthesia, and all efforts were made to minimize suffering.

2.2 Irradiation

The animals were randomly divided into six groups, n= 6 for each group. They were anesthetized by ketamine (0.05 mL/kg i.v.) before X-ray irradiation. The non-irradiated group was the control, while five treatment groups were irradiated to 2, 5, 8, 11 and 14-Gy of total body single dose on an 8-MV X-ray linac (Elekta Synergy Platform, ELEKTA Ltd, Sweden), at a dose rate of 255 cGy/min. The linac was standardized by CRS-3D tank system (MED-TEC USA) before irradiation. Radiation field was calculated as 2ab/a+b (a: length, b: width). Source-axis distance was 100 cm; the source-surfaced distance (SSD) was 85 cm (a mini pig was deemed as 30 cm thick). The machine output was converted into the dose by Dm=DT/TPR×SADF×Sc.P (TPR: tissue phantom ratio; SADF: source-axis distance factor; Sc.P: total scattering calibration factor; Dm: monitor unity dose; DT: tumor dose).

2.3 [¹⁸F]-FDG-PET/CT scan

Minipigs were scanned by FDG-PET/CT (Discovery-LS PET/CT, GE, USA). All images and data were processed by a Xeleris Workstation system (GE, USA). The quantization of FDG uptake was analyzed by regions of interest (ROIs). Based on the ROIs placed on liver, uptake was calculated as percentage of the irradiation dose (% ID). The normalization and correction of data were completed by computer automatically and presented by maximum standardized uptake value (SUV). The color of picture was determined by SUV and presented by a color bar. The following measures were taken to ensure the objectivity and accuracy of SUV and images: 1) we injected FDG tracer at a fixed dose range (0.40×10⁷–0.48×10⁷ Bq/kg/pig on Day 7 after TBI; 2) all pigs were fasted at least 8 hours before PET/CT scan; 3) the liver of each pig was scanned in three neighboring layers and collected the maximum value of SUV in each layer for statistical analysis; and 4) the minipigs were scanned 1 hour after FDG injection.

2.4 Blood sampling and processing

For the assay of alanine aminotransferase (AST), aspartate aminotransferase (ALT) activity and ALT/AST ratio, venous blood was collected from the ear vein of pigs. The serum was separated, and the AST and ALT activities were tested using Automatic Chemistry Analyzer (DH-1680. Nanfen Corp. Nanjing, China). For the assay of International Normalized Ratio (INR), capillary blood was collected and tested, using CoaguChek XS instrument (CoaguChek XS. Roche Diagnostics GmbH, German). The finger stick and application of blood drop on the strips were performed according to laboratory procedure by the same trained personnel during the whole study. Three experienced biomedical laboratory scientists performed the testing on the CoaguChek XS and INRatio.

2.5 Pathologic analysis

The animals were euthanized and sacrificed by haemospasia one week post-TBI. Livers were rapidly removed and fixed in 10% neutral buffered formalin and embedded in paraffin. Histological sections were cut into 4-μm sections and stained with Hematoxylin and eosin (H&E). To observe the changes of pathological type and the severity in different radiation doses and observation times, the severity of pathology grading of hepatic lesions was quantified by conventional Histology Activity Index (HAI) scores [6]. Pathological type was observed and the range was quantified by the following criteria: (1) normal livers were scored 1; (2) <1/4 of each high power field (HPF) showing pathologic change was scored as 2; (3) 1/4–1/2 each HPF showing pathologic change was scored as 3; (4) <1/2–3/4 each HPF showing pathologic change was scored as 4; and (5) <3/4–1 each HPF showing pathologic change was scored as 5. In addition, we measured countable hepatocytes in three randomly-selected areas from each pig, at a magnification of × 200 per section, using the Image-Pro Plus 6.0 image analysis package (Media Cybernetics, USA).

2.6 Statistical analysis

Data are presented as means ± standard deviation (SD) as indicated. The ANOVA test was used for statistical analysis of SUV, CT values and liver function. Post hoc multiple comparisons in each group were measured with the least significant difference (LSD) test or the Games-Howell test as determined by a homogeneity of variance test. Pathology data was measured using a nonparametric K independent Samples test. P value less than 0.05 was considered statistically significant (two-tailed). The correlation was analyzed based on Pearson (measurement data) and Spearman test (numeration data). All data were processed by Statistical Product and Service Solutions 13.0 (SPSS 13.0) software.

3.1 SUV and CT values of liver tissue

As shown in Fig. 1, the irradiation dose affected the hepatic SUV significantly (p < 0.05). In contrast to hepatic SUV, CT values of 0 Gy–14 Gy (p>0.05) did not differ significantly (Fig. 2). The SUVs in the 2-Gy (1.95± 0.49), 5-Gy (3.64±0.64), 8-Gy (4.42 ± 1.88) and 11-Gy (5.85 ±1.63) groups were higher than those in the control group (1.16 ± 0.14) (2, 5, 8- and 11-Gy vs. control, p < 0.05), while the SUVs in the 14-Gy (0.63± 0.10) group is lower than that in the control group (14-Gy vs. control, p < 0.05). Furthermore, the SUVs of 2 Gy–11 Gy increased with the dose (2-Gy vs. 5-Gy, 5-Gy vs. 8-Gy and 8-Gy vs.11-Gy, p < 0.05).

Fig. 1.Full-screen View

(Color online) FDG-PET/CT images and the tendency of liver (cross section+ coronal section) on 1 week after TBI respectively. The color scale indicates FDG uptake and is determined by SUV, the scale of red to blue represents SUV from 6 to < 1. A large contrast of FDG uptake in the irradiated groups can be seen. The error bars indicate SD.∗: p < 0.05, ^∗∗: p < 0.01(versus control); #: p < 0.05, ##: p < 0.01 (versus each other, analyzed by Post-Hoc Multiple Comparisons).

Fig. 2.Full-screen View

CT values in different radiation doses one week post-TBI. The error bars indicate SD.

3.2 Pathology findings

In Fig. 3, microscopic examination of pig liver sections of 2- and 5-Gy groups one week post-irradiation were compared to the control group. The structure of hepatic cord of the irradiation groups was still integrity, but the hepatic sinusoid around central veins of liver had an extension and congestion. In addition, hepatocytes presented pyknotic nucleus, cellular swelling and eosinophilic body. In contrast with the sections from 2- and 5-Gy groups, lesions were more severe in the 8-, 11- and 14-Gy groups one week post-TBI, including: 1) evidence of severe hepatic cord structure damages; 2) severe and obvious extension, congestion and hemorrhage in hepatic sinusoid and inner blood vessel of liver (localized or generalized); and 3) the necrosis of hepatocytes. Furthermore, specific changes were found in irradiation groups one week post-TBI (Table 1). Hemorrhage and necrosis were mainly in 8-, 11- and 14-Gy groups, and pyknotic nucleus, increasing of cellular swelling and eosinophilic body were the most distinctive change in 2- and 5-Gy groups. In short, the higher irradiation doses lead to more serious injury.

Fig. 3.Full-screen View

(Color online) Pathologic changes of liver tissues on 1 week after TBI (H&E 4/40×). Black arrows indicate the radiation-induced abnormality in liver tissue.

Pathologic score changed with the dose (Fig. 3). The HAI score showed a dose-dependent increase until 11-Gy and tended towards stability in 14-Gy while the peak level was recorded in 11-Gy and scored 4–5.

Figure 4 shows that the irradiation dose had significant effects on hepatocytes numbers (p < 0.05). Hepatocytes numbers of 2 Gy (492.86±10.05) and the control group (498.55±24.00) are about the same, while hepatocytes numbers in 5-Gy (447.72±11.75), 8-Gy (293.49±9.68), 11-Gy (303.77±18.11), 14-Gy (292.60±6.13) are lower than those in control group (p < 0.05). However, hepatocytes of 8, 11 and 14 Gy do not differ significantly (p > 0.05).

Fig. 4.Full-screen View

Hepatocytes numbers one week after TBI. Error bars indicate SD. ^∗: p < 0.05, ^∗∗: p < 0.01(versus control); #: p < 0.05, ##: p < 0.01(versus each other, analyzed by Post-Hoc Multiple Comparisons).

3.3 Liver function

Zymological test of liver mainly included serum ALT, AST and ALT/AST ratio (Table 2). Serum AST, ALT and AST/ALT of 2 Gy–14 Gy were of significant difference (p < 0.05). The AST, ALT values and AST/ALT in irradiation groups were higher than that of the control group (p < 0.05). AST values of 2 Gy–11 Gy were of the same level, but AST value of 14 Gy increased abruptly. ALT values of the irradiation groups increased and showed a fluctuation change trend from 2 Gy to 14 Gy. AST/ALT of the irradiation groups increased, tending stable from 2 Gy to 8 Gy, and further increased from 11 Gy to 14 Gy.

In Table 2, INRs of 2 Gy–8 Gy did not differ significantly from the control (p>0.05), but INRs of 11 and 14 Gy were significantly higher than the control. The dose did not affect hepatic serum albumin (p > 0.05). TBILs of 0 Gy–8 Gy were of the same level (p > 0.05), but TBIL of 11 and 14 Gy were significantly higher than the control.

The purpose of correlation analysis was to evaluated whether these indicators can be used to reflect the hepatic injury induced by radiation using the pathological findings as gold standard (Table 3). Our results showed that hepatic SUV is positively correlated with HAI score in the dose range of 2 Gy–11 Gy, and the correlation coefficient and p value is 0.573 and 0.022, respectively (Figure 5). Whereas, AST, ALT, AST/ALT, INR, SALB, TBIL showed no correlation with HAI score (p > 0.05).

Fig. 5.Full-screen View

The correlation between SUVmax and pathological scores.

3.4 Acute lesions of liver tissue after TBI exhibited a dose-dependent increase

TBI of normal liver tissue with a single dose of 2 Gy–14 Gy caused different degrees of hepatic lesions in pathology. In general, low doses (2 Gy–5 Gy) irradiation induced minor changes in hepatic histology at a microscopic level - pyknotic nucleus, increasing of cellular swelling, eosinophilic body and the mild loss of hepatocytes, while high doses (8 Gy–14 Gy) irradiation induced a spectrum of the pathological features with microscopic evidence of hepatic lesions–the marked loss of hepatocytes, necrosis and hemorrhage (generalized or localized). Combining the results of pathologic damage and HAI score, acute lesions of liver exhibited a dose-dependent increase one week after TBI. These findings are compatible with prior investigations [7, 8].

Perhaps in no field of radiation biology are opinions as greatly divergent as in assessing the response of the liver to irradiation. The liver was once regarded as a relatively radioresistant organ [5, 9-11]. It was reported that the liver could tolerate a considerable amount of radiation without significant damage (>50 Gy). However, later studies suggested that the tolerance irradiation doses of liver might be overestimated. Jirtle et al. [12] and Antoch et al. [13] revealed that fractionated doses greater than 20 Gy can result in severe radiation-induced liver disease (threshold dose for whole-liver irradiation without chemotherapy is 20 Gy–30 Gy). However, all the above controversial views were based on the results of chronic or subacute radiation toxicity of liver- the morbidity of radiation-induced liver disease (RILD) and fibrosis.

The RILD, namely, the changes that arise as early as approximately two weeks after hepatic irradiation, which are accompanied by a loss of parenchymal hepatocytes and the distortion of the lobular architecture [3]. Although hallmark of RILD is veno-occlusive disease and characterized by thrombosis of the central veins [3], our findings did not show any evidences of this. We assume that the unobserved evidences maybe result from the short time-course. In contrast, our findings in pathology revealed the marked loss of hepatocytes and severe damage one week after a single dose of 8 Gy–14 Gy (8 Gy in one fraction= approximately 10-Gy in 5 fractions, 11 Gy in one fraction = approximately 14 Gy in 7 fractions, 14 Gy in one fraction = approximately 18 Gy in 9 fractions, converted by Multitarget Model portion). These findings proved that the dose range of 8 Gy–14 Gy TBI was already sufficient to lead to severe liver damage. Knowledge of this acute radiation response on normal liver may be helpful for assessing the tolerant dose of liver given before irradiation and potentially led to more considerable attention on acute radiation toxicity of liver than ever before.

3.5 [¹⁸F]-FDG-PET/CT can be used to assess acute radiation injury of liver attributed to the close correlation with pathological findings

Currently, a controversy existed on noninvasive monitoring the response of liver tissue to irradiation using CT [14-17]. However, this controversy is based on the long-termed observation and prognostic purpose. In this study, the finding in CT value revealed that it is useless in monitoring the acute response of liver after TBI. It was reported showed the usefulness of FDG/PET-CT in accurate monitoring intraoperative radiation therapy in liver tissue [13]. However, a concern in such application to irradiated liver tissue is based on subacute change. No evidences exist concerning the acute effect of irradiation on glucose metabolism using FDG in normal liver tissue.

In our study, we found that hepatic FDG uptakes showed the similar trend with pathological score in the dose range of 2 Gy–11 Gy. We also found that hepatic FDG uptake in the 14 Gy group decreased abruptly and even lower than control. The results from our previous study may be helpful to explain the abrupt decrease of FDG uptake in 14 Gy [18]. In that study, the early reduction of marrow FDG uptake in the 11 and 14 Gy was found, and explained by the trade-off between the infiltration of the inflammatory cells and the relatively decreased cellularity. However, the level of hepatocytes loss showed no differences among the 8, 11 and 14 Gy groups even though the severe hepatocytes loss in the three groups was observed. Because the radiosensitivity and ability of tissue regeneration in liver was different from those of bone marrow, we supposed that the ability of liver regeneration may be involved (at least partially) besides the trade-off between decreased cullularity and pathological injury.

Correlation analysis showed hepatic SUV is closely correlated with pathological injury degree in the dose range of 2 Gy–11 Gy. Therefore, this kind of correlation between liver FDG uptake and pathology could make FDG PET diagnosis useful in assessing and monitoring the acute response of liver tissue to irradiation. We have no evidence on how long the correlation of FDG uptake with pahtology in irradiated liver will last and the relationship between FDG uptake with subacute and chronic toxicity of liver tissue. Therefore, careful sequential FDG PET evaluation of normal liver at various time points after irradiation will be needed to determine the significance of these preclinical findings.

3.6 Liver function may be inadequate to assess acute response of liver after TBI

Serum levels of AST and ALT increase greatly in hepatic dysfunction. The two indicators are used for RT-induced liver injury analysis [3]. Because the contents of AST are higher in heart, brain, and skeletal muscle [19], we used AST/ALT ratio to clarify whether the increase of AST level actually represented the TBI-induced hepatic damage or resulted from other TBI-induced organs injury. The marked increases of serum AST and ALT were observed in 11 and 14 Gy groups, and AST/ALT values in the two groups were higher than in control. Thus the increase of AST/ALT values certified the marked increase of AST and ALT in 11 and 14-Gy actually result from TBI-induced liver injury. However, the level of AST showed no differences in the dose range of 2 Gy–11 Gy, even though AST level of 2 Gy–11 Gy were greater than the control group (the same trend was shown in AST/ALT ratio). In addition, the fluctuated trend can be observed on ALT level. Therefore, the findings in zymology suggested limited diagnostic value and poor specificity for assessing acute responses of liver attributed to the irregular and fluctuated trend.

In addition, the INR and TBIL showed a slightly increase just in high dose TBI (8 Gy–14 Gy). As the important indicators, TBIL and INR were involved in the Child-Pugh scoring system to assess severity of liver disease. However, the highest level TBIL in our study is only 1.96±0.10. It is scored 1 in Child-Pugh scoring system [3]. Combing the pathologic result, therefore, they can not reflect the actual severity of radiation-induced liver injury.

3.7 The clinical and physiological impact from our findings

A well-established approach for monitoring radiation injury of liver is essential for therapeutic purpose (e.g radiotherapy for tumor control) and accidental overexposure (nuclear accident) [1]. Therefore, many studies devoted to assessing the radiation responses of the liver and its monitoring agents given after exposure. Although pathologic findings showed the evidences that TBI can cause different severity of liver damage within one week, several reasons inhibited their application for diagnostic purpose. First, the pathological procedures are associated with health hazards in patients who frequently have a poor performance status. Second, the abnormal tissues may occasionally be missed. It is therefore likely that FDG-PET/CT could greatly expand our ability to noninvasively and quickly assess and monitor radiation-induced organ toxicity.

Our results showed liver FDG uptake was correlated with pathologic findings. It suggested that [¹⁸F]-FDG-PET/CT is effective and precise for assessing the acute radiation response of liver for noninvasive diagnosis. In addition, PET/CT can acquire the data from whole body including a lot of organs and tissues. Therefore, our results also suggest the use of PET/CT to evaluate other organs as a safe and noninvasive approach because multi-organ monitoring is the unique advantage of PET/CT compared with other approaches for radiation-induced injury. Because the initial prospective study design from which these data were obtained was focused on acute response on liver after TBI, our study was presented with a number of challenges including: 1) the lack of long-termed observation to reveal whether the findings in acute period correlated with subacute changes (that occur at a month) of RILD or chronic changes (that occur at 6 months) of radiation fibrosis; and 2) the study is chiefly descriptive and there is not much study of how radiation is causing the observed events. Therefore, the further experimental purposes should be focused on the large-scale multi-centre clinic trial, and find the more detail in the relationship between the acute and subacute and chronic changes.