The feasibility of low-dose CT perfusion imaging in gastric cancer

Abstract

PURPOSE:

To investigate feasibility of applying low-dose CT perfusion imaging (CTPI) to diagnose gastric cancer.

MATERIALS AND METHODS:

Twenty patients with gastric cancer confirmed by endoscopic biopsy were undergone routine dose (120 kV, 100 mA) and low-dose (120 kV, 50 mA) CTPI examination, respectively. The original data were processed by body perfusion software, and the perfusion parameters values including blood flow (BF), blood volume (BV) and permeability surface (PS) of gastric cancer were measured. Statistical data analyses including paired-samples t test, Pearson correlation analysis and Bland-Altman consistency test were used to compare the perfusion parameters values between the routine dose and low-dose CTPI examinations. Radiation dosage, which the patients received during two CTPI examinations, was also calculated and compared.

RESULTS:

There were no statistical differences in the BF, BV and PS values between routine dose group and low-dose group (P > 0.05), and there were significant correlation in the BF, BV and PS values between two groups (P < 0.01). The consistency of BF and BV values between the two groups was preferable to that of PS value. The radiation dosage of the low-dose group was much less than that of routine dose group, and the CTDIvol and DLP values of low-dose CTPI were decreased by 50%, respectively.

CONCLUSION:

The parameters BF and BV values may play a valuable role in the diagnosis and assessment of gastric cancer in low-dose CTPI examination.

Keywords

Low-dose computed tomography perfusion imaging gastric cancer

1 Introduction

The incidence and mortality of gastric cancer are ranked first among malignant tumors of the digestive tract in China [1], presenting a serious threat to human health. An imaging examination occupies a pivotal position in the diagnosis and assessment of gastric cancer. However, common examinations such as a gastrointestinal barium examination, conventional CT examination, as well as the invasive gastroscopy rely solely on morphological changes for judgement. Therefore, we attempt to improve the detection of gastric cancer from the perspective of hemodynamics. CT perfusion imaging (CTPI) is a functional imaging technique, that is, the tumor’s blood supply and vascular permeability can be reflected through measuring perfusion parameters of tumor tissue.

Currently, CT perfusion imaging has been widely applied and tested in the clinical practice to detect and diagnose diseases among many human organs including brain [2], head and neck [3], lung [4], liver [5], pancreas [6] and others. Perfusion CT has also been able to assess therapy response of gastric cancer after radiotherapy and chemotherapy [7]. When CTPI is used to predict and assess treatment effect of gastric cancer, multiple repeated CTPI examinations are needed; however, excessive radiation is harmful to patients. Therefore, low-dose CTPI program is urgently required. At present, low-dose CTPI technique has been widely used in chest examination [8, 9], and the use in the abdomen has been occasionally reported [10]. In this study, whether low-dose CTPI technique can be used in gastric cancer or not, this problem is explored through analyzing perfusion parameter values of gastric cancer between routine dose group and low-dose group.

2 Materials and methods

2.1 Patients

This study was approved by Ethics Committee and each subject signed “Informed consent about CT perfusion scan”. Twenty patients with gastric cancer confirmed by endoscopic biopsy in our hospital from November 2016 to February 2017 were recruited. Each patient had two CT scans with both routine and low dose, that is, they were undergone routine dose (120 kV, 100mAs) and low dose (120 kV, 50mAs) CTPI examination, respectively. Among them, including 13 males and 7 females with an age range of 39–75 years old (mean age, 57.8 years old). There were 14 cases of cardia cancer, 3 cases of gastric antral cancer and 3 cases of gastric corpus cancer in 20 cases of gastric cancer with a diameter range of 2.5–6.5 cm (mean diameter, 3.5 cm). Inclusion criteria for a patient: the diameter of gastric tumor was more than 2 cm, the patient could hold breath for 30 seconds and had not history of iodine allergy, a cubital vein of the patient could meet high flow rate requirement when CTPI examination was performed.

2.2 Perfusion CT study protocol

The procedures of preparations before CTPI examination have been described and reported in our previous publication [11]. In brief, all patients were first conventionally plain-scanned on a 64-slice spiral CT scanner (Siemens Somatom Sensation, Germany) with the following parameters: a slice thickness and spacing of 5 mm, tube voltage of 120 kV, tuber current of 100 mAs, a matrix of 512×512 pixels, and pitch of 0.984:1. Then, CTPI examination was performed according to the plain CT image, the slice containing the largest tumor area was selected as the center slice, and a total of 6 nearby slices including the center slice were selected every 3 cm on the Z-axis. These slices were scanned with a small field-of-view (SFOV) at a 4.8 mm slice thickness, tube voltage of 120 kV, tuber current of 100 mA (routine dose group) and tuber current of 50 mA (low-dose group). A 50 mL bolus of iopamidol (Iopamiro; Bracco; Shanghai, China) containing 300 mg of iodine per mL was injected using a power injector (MEORAO-Stellant, MEORAO Company, Germany) via an antecubital vein at a rate of 5 mL/s through an 18-gauge intravenous cannula. Scanning commenced at 7 s after contrast agent injection and the scan duration was 30 s.

2.3 Image interpretation

CTPI examination data were transmitted to a post-processing workstation (ADW4.0, Siemens Company), according to our previous study experience [11], an abdominal tumor perfusion protocol in the body CT perfusion 3.0 soft-ware (deconvolution method) was then used for data processing. The abdominal aorta was selected as the input artery to obtain CT perfusion pseudo-color images and perfusion parameters BF, BV and PS values. All perfusion parameter values were measured three times and the averages of the 3 measurements were shown. The measurement and statistical analysis of all parameter values were performed by the author alone so as to ensure the consistency of parameters values.

2.4 Data analysis

SPSSl7.0 software was used for the statistical data analysis. The differences of BF, BV and PS values of gastric cancer between routine dose CTPI group and low-dose CTPI group were analyzed by using paired-sample t test. The correlation of BF, BV and PS values between two groups was analyzed by Pearson statistical method. In addition, the Bland-Altman plot made by GraphPad Prism 6 software was used to further analyze the consistency of perfusion parameter values between the two groups. In data analysis, P < 0.05 was used to indicate that the difference of the comparison was statistically significant.

3 Results

There were no statistical differences in BF, BV and PS values of gastric cancer between routine dose group and low-dose group (P > 0.05), respectively (Table 1).

Table 1
Comparison of CT perfusion parameters values of gastric cancer between routine dose group and low-dose group

Group n BF BV PS

(ml/100 g/min) (ml/100 g) (ml/100 g/min)

Routine dose group 20 100.73±17.77 18.36±5.27 34.33±17.76

Low-dose group 20 99.49±22.48 16.52±5.69 32.74±18.08

t value 0.56 1.94 0.46

P value 0.58 0.07 0.65

Group	n	BF	BV	PS
Routine dose group	20	100.73±17.77	18.36±5.27	34.33±17.76
Low-dose group	20	99.49±22.48	16.52±5.69	32.74±18.08
t value		0.56	1.94	0.46
P value		0.58	0.07	0.65

Note: All parameters values were expressed as mean±standard deviation ( $\bar{x}$ ±s), and presented normal distribution and homogeneity of variance; paired-sample t test was used to compare two groups, and P < 0.05 indicated that the difference was statistically significant. Abbreviations: blood flow (BF), blood volume (BV) and permeability surface (PS).

Correlation analysis was performed by Pearson statistical software for BF, BV and PS values of gastric cancer between routine dose group and low-dose group. The results showed that there was significant correlation for BF value between two groups (r = 0.906, P < 0.001) (Fig. 1A); there was significant correlation for BV value between two groups (r = 0.681, P = 0.001) (Fig. 2A); as well, there was significant correlation for PS value between two groups (r = 0.632, P = 0.003) (Fig. 3A).

Fig.1

f Figure 1A showed there was significant correlation for BF value between routine dose group and low-dose group (r = 0.906, P = 0.000); Bland-Altman plot showed BF value consistency analysis between two groups. 100% (20/20) of points in BF plot were within 95% consistency limit (–20.53, 18.04) of differences (dotted line), absolute value of the maximum difference was 17.71, and mean of the differences was –1.24(solid line) (Fig. 1B). Abbreviation: blood flow (BF).

Fig.2

Figure 2A showed there was significant correlation for BV value between routine dose group and low-dose group (r = 0.681, P = 0.001); Bland-Altman plot showed BV value consistency analysis between two groups. 5% (1/20) of points in BV plot wasn’t within 95% consistency limit (–6.46, 10.14) of differences (dotted line), absolute value of the maximum difference within 95% consistency limit was 6.09, and mean of the differences was 1.84 (solid line) (Fig. 2B). Abbreviation: blood volume (BV).

Fig.3

Figure 3A showed there was significant correlation for PS value between routine dose group and low-dose group (r = 0.632, P = 0.003); Bland-Altman plot showed PS value consistency analysis between two groups. 15% (3/20) of points in PS plot weren’t within 95% consistency limit (–28.55, 31.74) of differences(dotted line), absolute value of the maximum difference within 95% consistency limit was 22.28, and mean of the differences was 1.60(solid line)(Fig. 3B). Abbreviation: permeability surface (PS).

Consistency analysis was performed by Bland-Altman statistical software for BF, BV and PS values of gastric cancer between routine dose group and low-dose group. The results of Bland-Altman plots showed that 100% (20/20) of points in BF plot were within 95% consistency limit (–20.53, 18.04) of differences, absolute value of the maximum difference was 17.71, and mean of the differences was –1.24 (Fig. 1B); 5% (1/20) of points in BV plot wasn’t within 95% consistency limit (–6.46, 10.14) of differences, absolute value of the maximum difference within 95% consistency limit was 6.09, and mean of the differences was 1.84 (Fig. 2B); 15% (3/20) of points in PS plot weren’t within 95% consistency limit (–28.55, 31.74) of differences, absolute value of the maximum difference within 95% consistency limit was 22.28, and mean of the differences was 1.60 (Fig. 3B).

The images of routine dose CT perfusion examination were all clear, and the gastric tumors were showed in red and yellow in the BF, BV and PS images (Fig. 4). As well, the images of low dose CT perfusion examination were all clear, and the gastric tumors were showed in red and yellow or blue in the BF, BV and PS images (Fig. 5).

Fig.4

Routine dose CTPI images with gastric cardia cancer. A: An original image showed the tumor in the cardia (arrow). B-D: They were CT perfusion BF, BV and PS images, respectively. In the BF and PS images, the tumor area was shown in red; In the BV image, the tumor area was shown in red and yellow. Abbreviations: CT perfusion imaging (CTPI), blood flow (BF), blood volume (BV) and permeability surface (PS).

Fig.5

Low-dose CTPI images with gastric cardia cancer, it was the same case as Fig. 4. A: An original image showed the tumor in the cardia (arrow). B-D: They were CT perfusion BF, BV and PS images, respectively. In the BF, BV images, the tumor area was shown in red and yellow; In the PS image, the tumor area was shown in blue. Abbreviations: CT perfusion imaging (CTPI), blood flow (BF),blood volume (BV) and permeability surface (PS).

According to CTDIvol and DLP data provided by CIPI scanning plan, the CTDIvol and DLP values of routine dose CTPI examination were 202.20mGy and 582.34 mGycm, respectively (Fig. 6). As well, the CTDIvol and DLP values of low-dose CTPI examination were 101.10mGy and 291.17mGycm, respectively (Fig. 7). It showed that the CTDIvol and DLP values of low-dose CTPI examination were reduced to 50%, respectively.

Fig.6

Figure 6 showed the CTDIvol and DLP values of routine dose CTPI examination (120 kV, 100mAs) were 202.20mGy and 582.34 mGycm, respectively.

Fig.7

Figure 7 showed the CTDIvol and DLP values of low-dose CTPI examination (120 kV, 50mAs) were 101.10mGy and 291.17 mGycm, respectively. Abbreviation: CT perfusion imaging (CTPI).

4 Discussion

In general, CT perfusion scan generated relatively large radiation dose, therefore it was necessary to correctly select scan method and optimize parameters for low radiation dose. After stomach cavity was filled with water, there was density difference between tumor tissue and water inside stomach cavity and fat outside stomach cavity; based on this condition, it was possible for us to use low-dose CTPI technique for a stomach tumor. In this study, we decreased radiation caused by CTPI examination through low tube current (50 mA). The results showed that there were no statistical differences in BF, BV and PS values of gastric cancer between routine dose group and low-dose group (P > 0.05, respectively), and there were significant correlations for BF, BV and PS values of gastric cancer between routine dose group and low-dose group (P < 0.01, respectively).

Bland-Altman analysis was first proposed by Bland JM and Altman DG in 1986 [12]. Its basic idea was to directly reflect consistency limit of two measurement results in a graphical manner. If distribution of differences followed a normal distribution, 95% consistency limit of differences was between $\bar{x}$ –1.96 SD and $\bar{x}$ +1.96 SD [13]. If all differences of two measurement results were within 95% consistency limit of Bland-Altman plot, we believed that the two scan methods had relatively good consistency and could be interchangeably used [14]. In this study, the results showed that 100% (20/20) of points in BF Bland-Altman plot were within 95% consistency limit of differences, therefore, we believed that the consistency of BF values between conventional CTPI and low-dose CTPI was relatively good; As well, 5% (1/20) of points in BV plot wasn’t within 95% consistency limit of differences, both absolute value of the maximum difference and mean of the differences were small, the difference was acceptable by clinic. So we also believed that the consistency of BV values between two scan methods was good; 15% (3/20) of points in PS plot weren’t within 95% consistency limit of differences, absolute value of the maximum difference was 22.28. The difference of PS value between two groups was relatively large and wasn’t accepted by clinic; therefore, we believed that the consistency of PS value between the two scan methods was unsatisfactory.

The International Commission on Radiological Protection believes that an increase of 1 mSv forX-ray exposure will increase malignant tumor incidence of 5/100,000 [15]. As more and more people are concerned about the danger cause of radiation-caused cancer, application of low-dose CT examination has become a hotspot at present. Naidich et al. [16] proposed concept of low-dose CT at first in 1990, that is, when tube current was decreased, radiation dose was decreased at the same time good image quality was kept to meet diagnostic demands for radiologists. Decrease of tube current is the most straightforward and handy method to reduce CTPI radiation. Because CTPI technology was a quantitative measurement method for perfusion parameters, we properly neglected image quality, accurate and stable perfusion parameters data could also be obtained. Watanabe et al. [9] carried out a liver CTPI study with conventional dose (120 kV, 100 mA) and ultra-low-dose (120 kV, 20 mA), the results showed that there was no difference in all CT perfusion parameters between the two groups (P > 0.05), while radiation dose in ultra-low-dose group was only 20% of that in conventional dose group. CTPI was performed by Wang et al. [17] in 34 healthy volunteers randomly divided into three groups with different applications of tube current, including a conventional dose group, a median-dose group and a low-dose group, the results showed that there were no significant differences between the parameters of the three groups, and the effective dosage of conventional, median and low-dose liver CTPI were 19.62 mSv, 12.61 mSv, and 7.01 mSv, respectively. The radiation dosage of low-dose liver CTPI was reduced to 64.27% compared with that of the conventional group. Whole pancreas CT perfusion scan was underwent by Li et al. [18] in Sixty-one patients divided into two groups (group A:70 kV, 120 mA; group B: 80 kV, 100 mA), the results showed that there was no significant difference on CT perfusion parameters between group A and group B, and the effective radiation dose of the whole pancreas CT perfusion of group A and group B were 3.60 and 4.88 mSv, respectively. In this study, we performed CTPI examinations in twenty patients with gastric cancer divided into two groups (routine dose group: 120 kV, 100 mA; Low-dose group: 120 kV, 50 mA), we found that there were no statistical difference in the BF, BV and PS values between two groups, while radiation dose of low-dose CTPI was reduced to 50% compared with that of the routine dose group.

Our study had several limitations. Firstly, it was difficult for CT perfusion examination to detect small tumor and ulcerative gastric cancer, our study focused on gastric cancer which the diameter of the mass ranged from 2.5 cm to 6.5 cm. Secondly, the artifact due to respiration was minimized but not avoided, which may limit the quality of images and interfere with the results of perfusion parameters values. Finally, the cases of 20 patients with gastric cancer recruited in this study had a limitation of relatively small study group.

In summary, we found that applying low-dose CTPI could effectively reduce radiation dose, and perfusion parameters BF and BV values have great value in quantitative analysis to gastric cancer from the perspective of hemodynamics.

Conflict of interest statement

We declare that we have no conflict of interest.

Footnotes

Acknowledgments

This study was supported by the grants from the science foundation of Wuxi Health and Family Planning Commission (MS201643) and Wuxi Hospital Management Center (YGM123) in China. We would like to acknowledge Professor Hu for his technical help.

References

Yan

S.Y.

, Hu

, Fan

J.G.

, et al., Clinicopathologic significance of HER-2/neu protein expression and gene amplification in gastric carcinoma, World Journal of Gastroenterology: WJG 17(11) (2011), 1501–1506.

Tian

, Zeng

, Zhang

, et al., Robust low-dose dynamic cerebral perfusion CT image reconstruction via coupled dictionary learning scheme, Journal of X-Ray Science and Technology 24(6) (2016), 837–853.

Nesteruk

, Lang

, Veit-Haibach

, et al., Tumor stage, tumor site and HPV dependent correlation of perfusion CT parameters and [18F]-FDG uptake in head and neck squamous cell carcinoma, Radiotherapy and Oncology 117(1) (2015), 125–131.

Gibaldi

, Barone

, Gavelli

, et al., Effects of guided random sampling of TCCs on blood flow values in CT perfusion studies of lung tumors, Academic Radiology 22(1) (2015), 58–69.

Liapi

, Mahesh

, Sahani

D.V.

Is CT perfusion ready for liver cancer treatment evaluation?

Journal of the American College of Radiology 12(1) (2015), 111–113.

Pogue

B.W.

, Elliott

, Samkoe

K.S.

, et al., Perfusion CT examined as a surrogate dosimetry tool to estimate verteporfin uptake in rabbit orthotopic pancreas cancer, Photodiagnosis and Photodynamic Therapy 12(3) (2015), 370–370.

Hansen

M.L.

, Fallentin

, Lauridsen

, et al., Computed tomography (CT) perfusion as an early predictive marker for treatment response to neoadjuvant chemotherapy in gastroesophageal junction cancer and gastric cancer-a prospective study, Plos One 9(5) (2014), 1–10.

Murphy

, So

, Lee

T.Y.

, et al., Low dose CT perfusion in acute ischemic stroke, Neuroradiology 56(12) (2014), 1055–1062.

Spira

, Gerlach

J.D.

, Spira

S.M.

, et al., Effect of scan time on perfusion and flow extraction product (K-Trans) measurements in lung cancer using low-dose volume perfusion CT (VPCT), Academic Radiology 19(1) (2012), 78–83.

10.

Watanabe

, Katada

, Gohkyu

, et al., Liver perfusion CT during hepatic arteriography for the hepatocellular carcinoma: Dose reduction and quantitative evaluation for normal- and ultralow-dose protoco, European Journal of Radiology 81(12) (2012), 3993–3997.

11.

Sun

Z.Q.

, Cheng

X.F.

, Ge

Y.X.

, et al., Role of CT perfusion imaging in patients with variously differentiated gastric adenocarcinoma, Journal of X-Ray Science and Technology 23(6) (2015), 737–744.

12.

Bland

J.M.

, Altman

D.G.

Statistical methods for assessing agreement between two methods of clinical measurement, Lancet 327(8476) (1986), 307–310.

13.

Giavarina

Understanding Bland Altman analysis, Biochemia Medica 25(2) (2015), 141–151.

14.

Tian

S.J.

, Lu

Y.G.

, Liu

, et al., Comparison of 2 available methods with Bland-Altman analysis for measuring intracompartmental pressure, The American Journal of Emergency Medicine 34(9) (2016), 1765–1771.

15.

G.F.

, Li

H.L.

, Han

Z.S.

, et al., Clinical application status of multislice spiral CT low dose scanning, Chinese Journal of Radiological Health 21(2) (2012), 246–247.

16.

Naidich

D.P.

, Marshall

C.H.

, Gribbjn

, et al., Low dose CT of the lungs: Preliminary observations, Radiology 6(175) (1990), 729–731.

17.

Wang

W.J.

, Zhong

, Hua

X.L.

, et al., Low-dose hepatic computed tomography perfusion imaging and its preliminary study, Journal of Digestive Diseases 12(3) (2011), 204–209.

18.

H.O.

, Sun

, Xu

Z.d.

, et al., Low-dose whole organ CT perfusion of the pancreas: Preliminary study, Abdominal Imaging 39(1) (2014), 40–47.

Group	n	BF	BV	PS
		(ml/100 g/min)	(ml/100 g)	(ml/100 g/min)
Routine dose group	20	100.73±17.77	18.36±5.27	34.33±17.76
Low-dose group	20	99.49±22.48	16.52±5.69	32.74±18.08
t value		0.56	1.94	0.46
P value		0.58	0.07	0.65