Department of Diagnostic Imaging, Oncological Radiotherapy and Hematology, A. Gemelli University Hospital Foundation IRCCS, Largo A. Gemelli 8, 00168, Rome, Italy
Department of Diagnostic Imaging, Oncological Radiotherapy and Hematology, A. Gemelli IRCCS Polyclinic Foundation - Catholic University of the Sacred Heart, Largo F. Vito 1, 00168, Rome, Italy
Published Date: 12/08/2021.
*Corresponding author: Andrea Contegiacomo, Department of Diagnostic Imaging, Oncological Radiotherapy and Hematology, A. Gemelli University Hospital Foundation IRCCS, Largo A. Gemelli 8, 00168, Rome, Italy
Purpose: Analysis of a preliminary experience with CT-guided percutaneous drainage of pelvic collections/abscesses (CTPD) performed using iterative reconstruction algorithms (IRA).
Methods: CTPD procedures performed with IRA (20 patients; Study group, SG) were compared with CTPD performed with a standard protocol (20 patients; Control group, CG); study period: September 2015 – July 2019. Mean procedural time, number of CT scans, Mean Computed Tomography Dose Index (CTDI) and Dose Length Product (DLP) were assessed in both groups. Technical success and complication rate were evaluated. Qualitative (readers agreement on a 4-point scale; Cohen’s k) and quantitative (assessment of contrast-to-noise ratio, CNR and signal-to-noise ratio, SNR) analysis of CT images was performed between SG and CG.
Results: Mean procedural time was similar between SG and CG (21.7 ± 6.2 vs 22.6 ± 5.3 minutes; p=0.62) as well as the median number of CT scans (8 vs 8.5; p=0.71). CTDI (3.1 ± 2.6 vs 19.6 ± 10.8; p<0.001) and DLP (258.1 ± 198.4 vs 2069.4 ± 1059.3; p<0.001) were lower in the SG. Technical success was achieved in all the procedures. No complications were observed.
Conclusion: The use of iterative reconstruction algorithms for CTPD of pelvic collections/abscesses is feasible and provides significant dose reduction, without increasing intraprocedural complications.
Interventional radiology plays a key role in the management of pelvic collections and abscesses representing the real alternative to conventional surgery. The use of computed tomography (CT) as guiding modality has already been described with excellent outcomes and well know advantages in comparison with surgery [1,2]. On the other hand, CT increases the biological risk associated with the exposure to ionizing radiations, especially in patients who undergo repeated CT scans . The development of iterative image reconstruction algorithms (IRA) allowed ionizing radiation dose reduction, maintaining a high image quality of CT diagnostic examinations [4,5]. At the best of our knowledge, there is a lack of evidence in the literature about IRA-based CT protocols for interventional applications in the abdominal district.
This study reports a preliminary experience in reducing ionizing radiation dose during CT-guided percutaneous drainage of pelvic collections/abscesses (CTPD) using an IRA-based CT protocol.
The study was approved by the Institutional Review Board. Ethic committee approval was waived due to the retrospective nature of the study. All the procedures were performed by the same Operator with 3-year experience at the beginning of the study (September 2015). An IRA-based protocol for CTPD procedures performed with conventional CT technique was introduced at our Institution in September 2017. All the consecutive CTPD performed between September 2015 and July 2019 on the same CT scanner were retrospectively evaluated. Procedures performed with fluoroscopic CT technique, those performed by other Operators and those performed on districts other than the pelvic one, were excluded from the present study (Figure 1). Pelvic district was defined as the anatomical region between the axial plane through the iliac crests and the axial plane through the external anal sphincter.
Procedures were enrolled in two different groups on the basis of the IRA-based technical arrangement over time: consecutive CTPD procedures performed between September 2017 and July 2019 were included in the Study group. Consecutive CTPD procedures performed between September 2015 and July 2017 were included in the Control group.
Technical parameters for Study and Control group are reported in Table 1; in detail, CT parameters of the Study group consisted in the reduction of the rotation time (0,4 seconds vs 0,6 seconds of the Control group), of the kilovolt peak (80 vs 100), of the milliampere range (0-350 vs 50-400) and in the incrementation of pitch (1,53 vs 1,375), ASIR percentage (80% vs 20%) and noise index (70.0 vs 30.0).
CTPD: Computed Tomography Guided Percutaneous Drainage
Figure 1: Study Design.
CT: Computed Tomography; GE: General Electrics; mm: millimeters; SFOV: scan field of view; s: seconds; kVp: kilovolt peak; mAs: milliampere per second; ASIR: Adaptive Statistical Iterative Reconstruction.
Procedural steps and patients’ follow-up
Patient position and percutaneous access points were determined on the basis of a previous diagnostic CT examination. A CT scan was performed with skin markers for the needle path planning. After local anesthesia, an introducer needle (18 Gauge) was advanced into the pelvic collection/abscess. A guidewire was looped within the collection/abscess and a drainage tube was finally positioned, with or without pre-dilatation of the percutaneous path. A final CT scan was performed to confirm the presence of the tube pigtail into the target collection and assess possible intra-procedural complications.
All patients were clinically monitored the days following the procedure and a CT control was performed when indicated, such as in case of onset of abdominal pain or fever; a CT control before discharge or tube removal was performed in all patients. All patients were monitored the days following the procedure and a CT control was performed when clinically indicated, such as in case of onset of abdominal pain or fever; a CT control before discharge or tube removal was performed in all patients as scheduled for our internal protocol.
Data collection and variables definition
Patient-related (age, sex, Body Mass Index), collection-related (underlying cause, structure and size), procedure-related (technical success, patient position, tube size and indwelling time, intraprocedural complications, ionizing radiation dose, procedural time, and number of CT acquisitions), and post- procedural (clinical success, early and late complications) data were collected and compared between the study and control group.
The maximum diameter of each collection/abscess was measured on the axial images of the CT diagnostic examination performed before the procedure. The pelvic collection/abscess structure was defined as simple if unilocular, otherwise it was defined as complex. Technical success was defined by the placement of the drainage tip within the collection at the end of the procedure. Minor and major complications were recorded according to the SIR quality improvement guidelines for percutaneous drainage/aspiration of abscess and fluid collections . Ionizing radiation dose was estimated through the mean value of Computed Tomography Dose Index volume (CTDIvol) and the Dose Length Product (DLP).
Procedural time was estimated on the CT images as follows: the time of the first image after the localization scout was considered the beginning of the procedure and the time of the last image of the final CT control scan was considered as the end of the procedure. The number of CT acquisitions was calculated for each procedure, without considering the initial localization scout. Procedural feasibility was assessed through the comparative analysis of procedural time and number of CT scans in the Study and the Control group.
Image quality assessment
Qualitative and quantitative analysis was performed for CT images of both groups (Figure 2). Two interventional radiology residents (attending the 4th and 3rd year respectively) blindly classified the quality of each CT examination according to a four-point scale (1=inadequate; 2=sufficient; 3=good; 4=excellent) for the qualitative analysis; readers agreement analysis was subsequently performed. Mean contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR) of the CT examinations of Study and Control group were calculated for quantitative analysis assessment; a region of interest (ROI) was placed in the piriform muscle to assess muscular attenuation value (Apm) and standard deviation (SDpm). Another ROI was placed in the subcutaneous fat as reference tissue to obtain the same parameters (Af; SDf). CNR and SNR were calculated as follows: CNR = (Apm – Af)/SDf | SNR =ROIpm/SDpm.
Figure 2: Image quality. Comparative view of procedures performed in study (a, b) and control (c, d) group.
(a): Female, 60-year-old patient; body mass index: 27,2; SNR=2,04; CNR= 6,24; Reader 1 score: 3; Reader 2 score: 3.
(b): Male, 77-year-old patient; body mass index: 20,4; SNR=3,04; CNR= 8,14; Reader 1 score: 4; Reader 2 score: 4.
(c): Female, 67-year-old patient; body mass index: 26,4; SNR=2,78; CNR=9,57; Reader 1 score: 4; Reader 2 score: 4.
(d): Male, 46-year-old patient; body mass index: 18.6; SNR=2,13; CNR=7,91; Reader 1 score: 3; Reader 2 score: 4.
F: Female, M: Male; mm: Millimeters; CT: Computed Tomography; mGy: Milligray; mGycm: Milligray-Centimeters *Data are presented as mean ± standard deviation (95% confidence interval)
A commercially available software (SPSS 22, IBM) was employed for statistical analysis. Continuous variables with normal distributions were reported as mean ± standard deviation (95% confidence interval was calculated) and Student’s t-test was employed to assess significant differences in case of normal distribution. For nominal and ordinal variables or for variables without normal distribution, Mann-Whitney “U” test was adopted for inferential statistics. A p-value < 0.05 was considered as statistically significant and all statistical test were 2-tailed. Cohen’s k was calculated to assess observers’ agreement. Level of agreement was defined as follows: k<0.20, slight agreement; k=0.20–0.40, fair agreement; k=0.41–0.60, moderate agreement; k=0.61–0.80, substantial agreement; k=0.81–1.0, almost perfect agreement.
BMI: Body Mass Index; mm: millimeters; G: Gauge; min: minutes
A total of 40 patients were included, 20 in the Study group (mean age: 62.1 ± 15.1; range: 37-97 years; 11 females) and 20 in the Control group (mean age: 63.1 ± 12.9; range: 40-86; 13 females) respectively. No differences were observed in age (p=0.81), sex distribution (p=0.75), and Body Mass Index (p=0.8) between the two groups; a comparative analysis of the main variables between the groups is reported in Table 2.
Technical success was achieved in all the procedures in both groups and no major or minor intraprocedural complications were observed. 3/20 (15%) patients of the Study group had minor complications after the procedure: in patient 3, tube malfunction was observed after 5 days; in patient 6, poor but continuous tribute from a 10 French (Fr) drain tube indicated substitution with a 12 Fr “two-way” tube after 43 days; similar management was observed in patient 8, in which a 10 Fr tube was replaced by a 14 Fr tube after 11 days. Patient 16 needed a re-intervention three days after the procedure for a considerable anastomotic leak.
The mean indwelling time of the tube was 15.6 ± 11 days and 12.6 ± 8.8 days for the Study group and the Control group, respectively (p=0.35). A descriptive analysis of the Study group variables is shown in Table 3. The mean procedural time was 21.7 ± 6.2 minutes in the Study group and 22.6 ± 5.3 minutes in the Control group (p=0.62). The median number of CT scans was 8 (range: 3-14) in the Study group and 8.5 (range: 4-21) in the Control group with a p value of 0.71. The mean CTDIvol value was lower in the Study group as compared to the Control group (3.1 ± 2.6mGy vs 19.6 ± 10.8 mGy; p<0.001), as well as the mean DLP values (258.1 ± 198.4 mGy-cm vs 2069.4 ± 1059.3 mGy-cm; p<0.001).
Image quality assessment
Qualitative analysis showed substantial agreement between the readers for the study group (Cohen’s k=0.74), with 36/40 observations scoring 3-4 points, and almost perfect agreement for the control group (Cohen’s k=0.82) with 38/40 observations scoring 3-4 points. CNR assessment showed a value of 9,05 ± 4,53 for study group and 8.21 ± 3.28 for control group respectively (p=0.51). SNR was 2.56 ± 1.45 and 2.58 ± 1.23 for study group and control group respectively, with a p value of 0.92.
The present study reports a preliminary experience with a conventional CT IRA-based protocol for the percutaneous management of pelvic collections and abscesses. The comparative analysis between the Study and Control group suggests that a conventional CT IRA-based approach with ionizing radiation dose reduction is feasible and safe. Qualitative images analysis showed high level of agreement between the readers and confirmed a good image quality for interventional purposes. In addition, both SNR and CNR mean values were maintained comparable to those of the control group supporting the concept that the discrimination of the structures surrounding the collection/abscess was adequate during the study group procedures. This aspect is confirmed by the evidence that both average procedural time and number of CT scans were comparable between the groups, and the procedural workflow (i.e. images interpretation during the procedure) was not impaired in the study group. Similarly, the IRA-based approach has also proven to be safe, as no major or minor complications occurred in the Study group.
The relatively small sample size, the absence of randomization, the retrospective nature and the absence of data on the effective dose impose to consider our results as preliminary; however, the great difference in radiation dose parameters values between the groups, and both the absence of intraprocedural complications and the high rate of technical success in the Study group, suggest that a conventional CT IRA-based approach is possible.
The frequent use of CT as routine diagnostic modality is becoming a relevant problem with regards to ionizing radiation exposure of the general population . The same concern occurs in the interventional field: interventionists play a pivotal role in modulating the use of ionizing radiations according to the underlying clinical problem, in each patient, but exposure to ionizing radiations is still extremely variable in the literature [8-10] due to the lack of standardized protocols . In addition, many patients require repeated post-procedural CT controls to monitor the therapeutic response , with a subsequent increased exposure to ionizing radiations.
Basing on this background, many Authors have already proposed low-dose protocols for thoracic [13,14] and skeletal interventions [15,16] where the constitutive high contrast resolution provided by the tissues is a lifeline for dose modulation. In particular, Chang et al.  proposed a CT acquisition protocol based on IRA for interventional chest procedures, concluding that the protocol was feasible, with a clear reduction of radiation exposure while preserving overall diagnostic acceptability, safety and precision. In the abdomen, the implementation of IRA-based protocols for CT interventions has never been evaluated. Lucey et al.  reported their experience about the use of a low-dose protocol for biopsy and drainage procedures performed under CT-guidance below the diaphragm, but their protocol wasn’t based on IRA. In their study, the reduction of milliamperage values (30 milliamperes) was the main technical measure in the study group in comparison with controls, in which the procedures were performed with a standard technique. However, although they achieved high percentages (87.5% – 97.6%) of technical success, the values of ionizing radiation dose, number of CT scans and average procedural time have not been reported, preventing the possibility of a comparison with the present study.
Preliminary data in the present study suggests that the modulation of ionizing radiation exposure using an IRA-based CT protocol is an achievable goal for the management of pelvic collections/abscesses; in particular, this technical measure maintains constant the technical success and do not increase the mean procedural time, the number of CT scans, and the complication rate, suggesting that the protocol is feasible and safe at the same time. Further analysis with prospective and randomized studies on larger cohorts of patients is required to confirm these preliminary promising results.
The article is not under consideration for publication elsewhere. Each author gave a valid contribution to and approved the present submission.
Declaration of Interest