The authors had full control of all the data and information presented in this manuscript. Written informed consent was obtained from all the patients involved in the study, and the entire study protocol was approved by the Ethics Committee.
Patients
Between January 2010 and December 2015, patients with a histological diagnosis of lung cancer were evaluated by a multidisciplinary lung tumour committee and selected by a radiologist with more than 10 years of experience in chest imaging for CTP performance. Figure 1 shows a flow chart of patients included in this study.
A total of 183 patients with a histological diagnosis of NSCLC were prospectively enrolled in our study. All of them underwent PCT before receiving any treatment (baseline study). A total of 53 patients who received treatment with cytotoxic CCT or concomitant CCT and RT underwent a second PCT study following the treatment (control study).
Inclusion criteria were histological diagnosis of NSCLC; absence of previous oncological treatment for this tumour; maximum tumour diameter larger than 2 cm; and first-line treatment with cytotoxic CCT, with or without associated RT. Exclusion criteria were PCT studies of poor technical quality because of movement artefacts, x-ray beam hardening, noise, or inadequate contrast enhancement; tumours difficult to separate from other lesions such as atelectasis, pneumonitis, or lymphangitis; patients who did not continue follow-up at our hospital; complete response or lesion less than 2 cm in diameter on the first PCT control after treatment; and presence of respiratory artefacts that could not be corrected by automatic movement correction algorithms.
The baseline study was performed taking into account that iodinated contrast had not been administered in the previous 24 h. The control study was conducted during the same examination as the first standard computed tomography control after treatment with CCT, performing the PCT first and subsequently the study of the entire chest. In patients who received RT, the exam was performed 2 weeks after the end of therapy to avoid inflammatory changes that could influence perfusion parameters.
A total of 53 patients participated in this study, including 42 men (79%) and 11 women (20%), with an age of 62.4 ± 9.9 years (mean ± standard deviation; range 35–79 years). The mean dose length product (DLP) for PCT was 463.8 ± 123.3 mGy cm (mean ± standard deviation; range 273–713) and the mean effective radiation dose was 6.40 ± 1.72 mSv (mean ± standard deviation; range 3.82–9.82).
A total of 16 patients were excluded: 6 patients because of poor technical quality (4 with respiratory artefacts; 2 patients with x-ray beam hardening artefacts due to tumour proximity to the superior vena cava and the bone for a tumour located in the pulmonary apex), 7 patients because of the absence of an identifiable tumour or an unmeasurable lesion in the second PCT study, 1 patient due to RT in the previous 2 weeks, and 2 patients in whom the entire tumour volume had not been included in the second study.
The electronic clinical histories were reviewed, and the following data were collected: age, sex, tumour histology, radiological stage at the time of diagnosis according to the TNM classification, type of CCT, association or non-association with concomitant RT, time elapsed in days since the basal PCT study was performed until the first day of CCT, time elapsed in days from the start of the treatment until the second PCT control, and volume and maximum diameter of the tumour in the axial plane before and after treatment.
Two radiologists with 14- and 12-year experience in chest radiology evaluated the first PCT control and classified the patient cases as complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD) using the RECIST-1.1 criteria [6], applied only to the lesion on which the PCT studies were performed. When there was disagreement, the two readers reached an agreement by consensus.
Patient preparation and PCT technique
Prior to the exam, the patient was trained to maintain apnoea throughout the study. When the patient was not able to breath-hold, she/he was trained to perform shallow respiration.
A dual-source equipment with 128 rows of detectors (Flash Definition®; Siemens, Forchheim, Germany) was used. Once the topogram was performed, a radiologist planned the study field, including the entire lesion along the z-axis. When the lesion was not clearly identified in the topogram, an unenhanced scan was performed to locate it.
Fifty mL of iodinated contrast was injected (Iopromide 300, Ultravist® Bayer Healthcare; Berlin, Germany) at 5 mL/s, followed by 50 mL of saline at the same rate.
The PCT acquisition was initiated 2 s after the injection of the contrast commenced, using the following parameters: 80 kVp and 90 mAs; 32 × 1.2-mm detector configuration; 0.33 s tube rotation time; 3 or 5 mm reconstructed image thickness, according to the tumour size; and B20f reconstruction kernel. The total time of the PCT study was always 45 s. The time interval between scans was 1.5 or 1 s, depending on the tumour size along the z-axis, which resulted in 30 to 45 scans in each tumour. The total length of the studies along the z-axis ranged from 4 to 15 cm.
Post-processing and image analysis
The data were transferred to a workstation (Multi-Modality Workplace®, Siemens, Forchheim, Germany) and processed using the Volume Perfusion Computed Tomography (VPCT) Body program. The PCT studies were post-processed and analysed by a senior chest radiologist with 14 years of experience who has received specific training in lung cancer perfusion post-processing, without knowledge of the results of the first standard computed tomography control after treatment with CCT.
First, the automatic motion and noise correction algorithms included in the VPCT Body software were applied. An arterial density-to-time curve was obtained by placing a region of interest in the thoracic aorta, where the unenhanced reference image was selected. The tumour volume was selected via manual segmentation, drawing the contours of the lesion in the axial, coronal, and sagittal planes, using a cut-off threshold of -50–150 UH, which permitted automatically excluding the normal pulmonary parenchyma, non-tumour vascular structures, and calcium, from the segmented volume.
The following perfusion parameters were calculated using a deconvolution model: BF, in mL/100 mL/min; BV, in mL/100 mL; PMB, in mL/100 mL/min, and MTT, in seconds. Each parameter was represented as a colour on parametric maps; numerical values were given as mean and standard deviation.
Statistical analysis
Quantile comparison graphs were used to evaluate if variables followed a normal distribution. All data were normally distributed except BF that was near-normally distributed.
The means of perfusion parameters between responders and no responders as well as between adenocarcinomas and epidermoid carcinomas were compared in baseline study using t test for independent samples. The means and standard deviations of the perfusion parameters were calculated for all of the patients in the baseline study and in the control study and were compared to one another using the t test for paired samples. The perfusion parameters were compared before and after the treatment, at the different response levels according to RECIST-1.1, for all histological subtypes, and in epidermoids and adenocarcinomas separately, using the t test for paired samples. The Pearson correlation coefficient was used to evaluate whether there was a relationship between changes in the perfusion parameters and changes in tumour volume or size.
The programs R (R Foundation for Statistical Computing, 2014, V 3.1.0) and SPSS® (IBM® SPSS® Statistics 20.0.0) were used for the statistical analysis. The level of statistical significance adopted was 0.05.