|Home | About | Journals | Submit | Contact Us | Français|
B M F Hendriks and M Kok have contributed equally
The purpose was to evaluate individually shaped contrast media (CM) delivery in CT pulmonary angiography (CTPA) for suspected pulmonary embolism (PE).
100 consecutive emergency patients with clinical suspicion of PE were evaluated. High-pitch CTPA was performed on a second-generation dual-source CT using the following parameters: 100kV, 200–250mAsref, rotation time 0.28s, 128×0.6mm col. and image reconstruction 1.0/0.8mm (B30f). Group 1 (n=50) then received a fixed CM bolus (300=mgIml−1, volume=90ml and flow rate=6mls−1); Group 2 (n=50) received a body weight-adapted CM bolus determined by dedicated contrast injection software. For analysis, groups were further subdivided into low-weight (40–75kg) and high-weight (76–117kg) groups. Technical image quality was graded using a four-point Likert scale (1=non-diagnostic; 2=diagnostic; 3=good and 4=excellent image quality) at the level of the pulmonary trunk and pulmonary arteries. Objective image quality analysis was performed by measuring contrast enhancement in Hounsfield units (HU) at the same levels. Attenuation levels>180HU were considered diagnostic.
All examinations were graded as diagnostic at each level. The individual minimum pulmonary attenuation was 184 and 270HU for Group 1 and 2, respectively. Mean attenuation was as follows: Group 1: 475±105HU (40–75kg) and 402±115HU (76–117kg), p<0.03. Group 2: 424±76HU (40–75kg) and 418±100HU (76–117kg), p=0.8. For Group 2, CM volumes were: 55±5ml (40–75kg) and 66±5ml (76–117kg), leading to 16–51% CM reduction.
Even under emergency conditions, individualized CM protocols can provide diagnostic and robust image quality in CTPA for PE with a substantial reduction of CM volume for lower weight patients, compared with a fixed CM protocol.
CM volume can substantially be reduced by using individualized CM protocols in CT angiography for PE without compromising the diagnostic image quality.
Pulmonary embolism (PE) is a major cause of morbidity and mortality, with acute PE being the most severe clinical presentation.1 CT pulmonary angiography (CTPA) is widely used to rule out or confirm the presence of PE and is considered the reference standard for this emergency indication.2,3 The improvement of scanning techniques with the invention of the multidetector-row CT has led to a high spatial and temporal resolution, better delineation of peripheral arteries and increased detection rate of subsegmental PE.3,4 In addition, CTPA has been shown to provide an alternative diagnosis in up to 70% of patients in whom PE is not confirmed.5 The diagnostic and clinical value of CTPA has already been firmly substantiated;5–8 thus, future research should focus on optimizing scan protocols and improving workflow.9,10
A considerable obstacle in current clinical practice is a substantial number of non-diagnostic scans, with reported rates ranging between 3.3% and 7.3%.11,12 However, for an emergency indication such as PE, the scan protocol should be expected to deliver robust and reliable results 24h a day, 7 days a week. According to current literature, a diagnostically enhanced CTPA requires an intra-arterial attenuation of at least 180Hounsfield units (HU), which allows for the visualization of pulmonary pathology in at least 90% of the patient population.11,13 The enhancement of pulmonary arteries depends on multiple factors: scan technique including tube settings (kV; mAs), scan duration, patient-related factors such as breathing (Valsalva) and body weight as well as contrast media (CM) administration parameters.11,14,15 Regarding the latter, both the amount of iodine injected per second [iodine delivery rate (IDR)] and the total amount of iodine [total iodine load (TIL)] are important factors in the enhancement of pulmonary arteries.5,14–16 One of the most influential patient-related factors is body weight; arterial enhancement and body weight have been found to correlate significantly, as well as enhancement and body mass index.14,15 Therefore, fixed CM protocols may result in attenuation values below the diagnostic level for heavier patients, whereas for skinny patients, attenuation may exceed required levels. Therefore, a solution could possibly be found in individualizing contrast application for each patient. This study aimed to prospectively evaluate individually shaped CM delivery in emergency CTPA for suspected PE.
A waiver of written informed consent was obtained from the local ethical committee (METC, ref. 14-4-198).
Inclusion criteria were: referral for CTPA for suspected PE; ≥18 years of age; and a glomerular filtration rate≥60mlmin−1×1.73m2. All patients were weighed by radiological technicians directly before CTPA; in case of immobile patients, weighing was sometimes obstructed, and the last known weight was recorded from the electronic patient files, when available. Patients were excluded when an alternative scan protocol (e.g. higher kilovoltage; kV) had to be used (n=11) or when body weight/injection data were incomplete (n=25). A total of 100 consecutive patients were included for data analysis.
All scans were performed on a second-generation dual-source multidetector-row CT scanner (Somatom® Definition Flash; Siemens Healthcare, Forchheim, Germany) using a high-pitch CT angiography (CTA) protocol (”Flash”) with a tube voltage of 100kV, tube current of 200–250mAsref (CareDose 4D™; Siemens) and pitch value of 2.6. A gantry rotation time of 0.28s and slice collimation of 128×0.6mm were applied. Image reconstruction was performed using a medium soft-tissue kernel (Siemens B30f) at a slice thickness of 1mm with an overlapping increment of 0.8mm.
Pre-warmed [37°C (99°F)] CM (Ultravist®, Iopromide 300mgIml−1; Bayer Healthcare, Berlin, Germany) was used for all patients. CM was administered using a programmable dual-head CT injector (Stellant®; Bayer) via an 18–20G intravenous injection catheter in the left or right antecubital vein.
The first 50 patients (Group 1) received CM injection according to our clinical standard PE protocol consisting of a CM phase of 75ml, as well as a mixed phase with total 30ml of 50% CM and 50% NaCl (TIL: 27gI). Flow rate was 6.0mls−1 (IDR: 1.8gIs−1). All CM injections were followed by a saline chaser of 30ml at the same flow rate.
The second 50 patients (Group 2) received individual CM injection protocols adjusted to body weight (kg) and scan duration (s), calculated by a dedicated contrast injection software (Certegra™ P3T; Bayer). This injection software is able to calculate CM volume and flow rate based on a non-linear relationship between patient weight and duration of the CT data acquisition.17,18 After the body weight of the patient and average scan duration for the specific acquisition are provided by the user, CM volume and flow rate will be adapted simultaneously in order to provide similar injection duration for each patient. All injections were followed by a saline chaser of 30ml at the same calculated flow rates.
For both groups, individual bolus timing in order to generate an individual start delay was performed using test bolus technique at the level of the pulmonary trunk: 20ml of CM was followed by 40ml of saline using the flow rates pre-calculated for the diagnostic scan. An overview of all injection parameters is listed in Table 1.
All images were immediately clinically evaluated by experienced radiologists. Information about the presence and location of PE was collected from the clinical reports.
Acquired data were analysed on a dedicated medical workstation (SyngoVia™; Siemens) using 1-mm transverse sections. Technical image quality of all scans was graded by an experienced research fellow (MK) who was blinded to the injection protocols, using a four-point Likert scale (1= non-diagnostic; 2= diagnostic; 3=good; and 4=excellent image quality) at the level of the left pulmonary artery (LPA) and right pulmonary artery (RPA) as well as at the level of lobular, segmental and subsegmental arteries. Technical image quality was determined as a combination of contrast enhancement, image noise and the presence of artefacts.
Objective image quality including contrast enhancement in HU, contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR) was measured by a research fellow (BH) who was blinded to the injection protocols. Regions of interest) were manually delineated on the axial thin slices at the level of the origin of the pulmonary trunk (PT) just cranial to the pulmonary valve and in both LPA and RPA (Figure 1). Sufficient image quality was defined as attenuation values >180HU and CNR values ≥10.19 SNR and CNR were defined according to the following equations:13
All injection parameters (volume, flow rate, peak flow rate and peak pressure) were continuously monitored by a data acquisition program (Certegra™ Informatics Platform; Bayer) and read out after each injection.
Statistical analysis was performed using SPSS v. 22.0 (IBM Corp., New York, NY; formerly SPSS Inc., Chicago, IL). For comparison between different body weights, patients were divided into two groups: 40–75kg (low weight) and 76–117kg (high weight). Groups were checked for normal distribution using the Kolmogorov–Smirnov test. Continuous variables were reported as the mean± standard deviation, and differences between groups were calculated using one-way analysis of variance with a Tukey test for post hoc comparisons. Categorical variables were reported as percentages, and the χ2 test was used to measure differences between groups. All p-values are two-sided, and a p-value below 0.05 was considered statistically significant.
Baseline characteristics of the study population are summarized in Table 2; no significant differences in baseline characteristics were found. For Group 1, mean age (years), body weight (kg) and body mass index (kgm−2) were 64±15 (range: 16–91), 75±17 (range: 49–109) and 26±5 (range: 18–39). For Group 2, these values were 63±16 (range: 21–92), 71±15 (range: 43–117) and 25±5 (range: 15–39), with associated p-values of 0.79, 0.26 and 0.25, respectively.
Mean flow rates in Group 2 were 5.2±0.4mls−1 for patients between 40–75kg and 6.1±0.4mls−1 for patients between 76–117kg, ranging from 4.2 to 7.6mls−1. CM volumes were calculated as between 42ml (40kg) and 76ml (117kg); IDR ranged from 1.26 to 2.28gIs−1 and TIL from 12.6 to 22.8g, respectively. Peak pressures for both groups ranged between 99±17psi and 122±33psi and did not differ significantly between Groups 1 and 2, with p>0.3 (Table 3). Maximal injection pressure of 325psi was never reached in any patient or for any flow rate.
TIL was 27g for Group 1 and ranged from 12.6 to 22.8g in Group 2. The mean TIL for Group 2 was 16±1.4g (40–75kg) and 20±1.4g (76–117kg), showing a significant CM dose reduction for all patients compared with Group 1, with p<0.01 (Table 3 and Figure 2). In patients who received an individualized bolus, CM volume could be reduced by 51%, 33% and 16% for the lightest patient of (44ml at 42kg), an average patient (60ml at 75kg) and the heaviest patient of Group 2 (76ml at 117kg), respectively.
For both weight groups, the iodine per kilogram body weight was significantly lower in Group 2, with p<0.01. Iodine per kilogram body weight in Group 1 was: 0.4gIkg−1 for 40–75kg and 0.3gIkg−1 for 76–117kg patients. For the same weight groups, values in Group 2 were 0.3 and 0.2gIkg−1, respectively (Table 3 and Figure 3).
The total incidence of PE in this study population was 18%. In Group 1, a total of 10 patients were diagnosed with PE, including 2 central, 3 lobar and 5 segmental emboli. Group 2 showed eight patients with PE; two patients had lobar PE, three patients had segmental PE and three had subsegmental PE. These are the most proximal points of the emboli; for an overview of the extent of disease in each patient, see Table 4.
For both groups, all scans were graded as diagnostic image quality at each anatomic level. The lowest grade was 2, and this grade was found in only one patient of each group. All other scans were graded as “good” or “excellent” at each anatomic level.
For Group 1, mean attenuation values in the PT and in the RPA and LPA were as follows (40–75kg vs 76–117kg): 475±105 vs 402±115HU (PT), 435±94 vs 369±104HU (RPA) and 426±101 vs 360±102HU (LPA). There was a significant difference in attenuation values between weight groups for PT, LPA and RPA, with all p-values<0.03 (Table 5). For Group 2, these attenuation values were as follows (40–75 vs 76–117kg): 424±76 vs 418±100HU (PT), 418±84 vs 408±90HU (RPA) and 413±81 vs 403±92HU (LPA). Within Group 2, no significant differences were found in attenuation values between both weight groups; all p-values>0.6 (Table 5).
On a group level, no significant differences in attenuation values were found between Group 1 and 2. However, attenuation values for Group 2 show a much smaller range (Figures 4 and and5).5). Both groups showed no non-diagnostic scans with a minimum mean pulmonary attenuation of 184HU for Group 1 and 270HU for Group 2.
CNR was acceptable for both groups, but better for Group 2, with values (low weight–high weight) of 16±6 and 15±5 vs 15±5 and 11±5. SNR values for Group 1 were 14±4 (low weight) and 11± 3 (high weight); for Group 2, these values were 15±5 and 13±3, respectively (Table 5).
Individualized CM bolus administration in CTPA provided equal-to-superior enhancement compared with a fixed protocol whilst using less CM volume. Moreover, greater consistency of vascular enhancement values—indicating a more reliable and robust protocol—was observed throughout Group 2, whereas the scans for Group 1 showed a steady decline in attenuation with increasing body weight. The observed homogeneity of attenuation throughout all patient weights is concordant with other studies using CM individualization in CTA; however, this research shows an additional reduction of CM volume as well.17,20
Some studies on body weight-adapted CM protocols for CTPA in selective, low-weight patient groups have been published: Holmquist et al used an 80-kV protocol with TIL of around 13g in a low-weight patient group, and Kristiansson et al reduced the TIL in a similar patient group even further to 9.6g. However, 8–12% of the examinations were regarded as suboptimal.21,22 In coronary CTA, a smaller standard deviation of intracoronary attenuation was reported when using a body weight-adapted injection protocol, indicating a smaller variation in attenuation values between individual patients.17,20
In CTPA, the lower HU limit is important as CM attenuation is typically even lower in the more peripheral arteries. A CTPA with arterial attenuation below 180HU may be considered to be of non-diagnostic quality to reliably detect subsegmental emboli as mean HU values for acute and chronic embolism are around 33 and 87HU, respectively.23 In this study, all patients met this crucial cut-off value of 180HU.
CNR was sufficient in both groups. CNR values above 10 are generally accepted in coronary CTA literature;19 however, there is currently no consensus regarding CNR values in CTPA. In this study, CNR decreased significantly with increasing patient weight in Group 1, whereas CNR remained constant in the group with individualized CM injections (Group 2). The observed decrease in CNR can be attributed to lower CM attenuation values in heavier patients in combination with higher image noise owing to higher body mass.19,24 A further improvement in CNR could be achieved by alteration of reconstruction methods, for example, with the implementation of iterative reconstruction.25–27 However, subjective image quality will not always be experienced as superior owing to possible loss of contrast between different densities.
All patients (up to 117kg) from the individualized CM group received less CM compared with standard group patients. The original fixed protocol in this study used a broad CM bolus (injection over 20s) with a relatively high flow rate (IDR of 1.8gIs−1), which had been chosen as a safe option. Problems regarding the realization of diagnostic quality in CTPA previously called for a safe “one size fits all” bolus, in order to prevent double scanning and bolus administration. Consequently, part of the CM reduction in the individualized group can be traced back to the comparison with a generous fixed CM bolus. Nevertheless, even attenuation in patients weighing 76–117kg from the individualized group was higher than that of patients of equal body weight in the standard group. The use of relatively smaller CM volumes in combination with higher flow rates indicates an inferior importance of CM volume and TIL compared with flow rate and IDR with respect to arterial enhancement, even under emergency conditions: IDR can be manipulated by adjusting either flow rate or iodine concentration of CM [IDR (gIs−1)=flow rate (mls−1)×iodine concentration (gIml−1)]. Some authors state that attenuation in CTA could be increased by using higher iodine concentrations of CM,28 whereas others state that increasing IDR—by increasing flow rate—is the most influencing factor regarding CM enhancement.16,29 Increasing flow rate might be preferable for optimizing CM application for CTA, since increasing iodine concentration of CM simultaneously increases CM viscosity. A phantom study conducted by Kok et al30 showed that low concentration, low viscosity and pre-warmed CM are beneficial for achieving low injection pressures.
Nowadays, contrast-induced nephropathy might not be as much of a concern as previously suspected.31,32 However, Lee et al33 found thyroid dysfunction in 22% of patients after a single iodinated CM injection, with urinary iodine excretion correlating to the TIL. According to a recent review, contrast-induced thyroid dysfunction mostly appears to be transient; however, there is a risk of complications.34 The precise relationship between CM volume and thyroid dysfunction remains unclear to this point; thus, administering smaller volumes of CM might prove to be a safe option and help to avoid delays in scheduling CTPA. Furthermore, smaller bolus volumes in other CTA protocols have been shown to reduce costs in medical clinics.35
With the current scan protocol, a tube voltage of 100kV was used. Already some advancements have been made using lower kV settings in CTA.25–27,36–38 A decrease in kV increases intra-arterial CM attenuation, because the higher photon energy at lower kV settings is closer to the iodine k-edge of keV.4,24,39,40 Many studies focused on low kV protocols; however, one must ensure to concordantly increase mAs to avoid excessive image noise.19,41,42 For future optimization, it may prove beneficial to further define the optimal IDR values and CM volumes per weight category that ensure a mean attenuation of 180–300HU for all kV settings regularly used in clinics, e.g. 120, 100 and 80kV.
For this study, only a dedicated scan protocol at 100kV was used. This resulted in exclusion of patients with body weights requiring higher kV protocols, e.g. 150kg body weight at 120kV.
The use of individualized CM protocols provides diagnostic and robust enhancement in emergency CTPA, as well as a substantial CM volume reduction in lower weight patients, compared with a fixed CM protocol.