1.
Study subjects
Patients with upper esophageal cancers or malignant thymomas scheduled for post-operative radiotherapy during the period June 2007 to May 2012 were screened and selected,
and then underwent HRCT and CTPI examinations for this study.
All patients received 3 dimensional conformal radiotherapy (3D-CRT).
Inclusion criteria were as follows: ①total irradiation dose 60 - 62Gy; ②radiation dose per time: 2Gy/2 days; ③course of radiotherapy approximately 8 weeks; ④ V20 ( defined as the percentage of pulmonary volume irradiated to >20 Gy,
i.e.
percentage of lung volume receiving more than 20 Gy dose in total lung volume) was limited between 20%-25%; ⑤no combined chemotherapy history; ⑥no pulmonary diseases such as chronic bronchitis,
emphysema,
pulmonary tuberculosis,
acute inflammation,
hilar lymph node metastasis,
involvement of pulmonary hilar vessels,
congenital or acquired disease of pulmonary blood vessels.
Therapy was completed as scheduled in all cases.
All selected patients underwent HRCT and CTPI examination pre- and post-radiotherapy,
namely on week 0,
week 4,
week 8,
and week 12.
This study was approved by the ethics review committee and all the patients have agreed to and signed the informed consent form.
51 patients were selected according to the criteria mentioned above.
All patients survived longer than 12 months (after radiotherapy) and were examined on follow-up visits.
2.
HRCT and 64-slice CTPI scanning
Using Siemens SOMATOM Sensation 64 spiral CT scanner,
lung scan was performed with the standard Body CTPI application.
In order to make sure that there was no motion during the CTPI procedure,
all patients were given respiratory training prior to the scans so that they were able to maintain about 18 seconds breath-hold.
CTPI was performed after acquiring the thoracic topogram and choosing the DynMulti protocol for Body Perfusion.
18 seconds end-inspiratory breath hold was used when the scan was performed simultaneously.
The CT scan region was limited to the radiotherapy region,
including main pulmonary artery or level 1 and level 2 branches when possible.
40 ml of Omnipaque (300 mg/ml) was injected intravenously via antecubital vein at 6.0 ml/s with 5 seconds delay.
Scanning parameters of CTPI: tube voltage 120kV,
tube current 80mAs,
rotation speed 0.33 S/R,
360 degree per rotation,
scan period 1.0 second,
recon 4 slices per period,
slice width and increment 7.2mm,
matrix 512×512,
and FOV 320 mm×320 mm.
HRCT scan was performed immediately after the CTPI scan with the following parameters: volume scan of the whole lung was completed within 5-7 seconds of quiet end-inspiratory breath hold,
with tube voltage of 120 kV and tube current of 100mAs.
8 mm and 2 mm slice thickness were used for reconstruction of chest CT and HRCT images,
respectively.
3.
Post-processing for the data of CTPI
All raw data acquired via Syngo CT workstation were processed with Body CTP software ,
using the maximum slope model to calculate perfusion parameters.
Drawing of the region of interest (ROI) was performed based on radiation field: ROIs were separately drawn in irradiated- and non-irradiated lung tissue with the consistent margin spacing of 5 mm with respect to pleura,
heart and edges of radiation field,
while avoiding great vessels.
Regional blood flow,
volume and permeability surface (rBF rBV,
and rPS) of regional lung tissue of irradiated and non-irradiated area were measured,
respectively.
Individual measurement differences,
measurement errors,
and other background factors were eliminated by introducing normalized regional perfusion values,
which were defined as the ratio of the perfusion values in irradiated area to the perfusion values in non-irradiated field,
namely rrBF,
rrBV,
rrPS,
etc..
The changes in normalized lung tissue perfusion values over the course of the study were monitored (including irradiated and non-irradiated tissue).
This comparison was performed on both ARILI and non-ARILI patients.
CTPI results and HRCT results were also evaluated.
4.
Criteria of clinical comprehensive diagnosis for RILI,
and evidences of positive judgments in HRCT and CTPI
ARILI is a clinical comprehensive diagnosis described in the RTOG standard,
and its severity is classified as 5 levels: Grade 0 - Grade 4[4].
Positive RILI results of any HRCT scan during the overall course of study (12 month after beginning of radiotherapy) are considered as diagnostic evidence of RILI.
This diagnosis is also supported by clinical manifestations,
laboratory examinations and follow-up to exclude the possibility of lung inflection.
The key goal of this study is to evaluate perfusion characteristics of the lung affected by RILI at the time when only half of overall radiation dose has been delivered (week 4).
Patients diagnosed with RILI on HRCT within 12 months were placed into RILI group,
and those with HRCT negative for RILI were placed into non-RILI group.
Positive HRCT was determined on the basis of Libshitz’s[12] study of x-ray and CT manifestations,
in which RILI was diagnosed if any one of the following was observed in the irradiated lung field: a) patchy exudation; b) patch of consolidation; c) atelectasis; d) chronic fibrosis.
In other words,
these radiologic features actually include ground-glass opacities,
patchy consolidation,
pleural reactions,
and lung fibrosis[13].
5.
Statistical analysis
All Data acquired were represented by means ± standard deviations (x±s) and analyzed using SAS (version 8.01).
Data acquired before irradiation (week 0) was compared to that acquired after irradiation (week 4-12) using t-test to find the difference in Mean values.
One-way ANOVA (F-test) was used to find the difference in mean values of data from the same patient obtained at different acquisition times and Dunnett-t test was used to compare in different groups.
The results of diagnostic accuracy by CTPI and HRCT were compared using Chi-square test (χ2) and Fisher' exact method.
A P <0.05 was considered indicative of significant difference.