Computed tomography pulmonary angiography using high-pitch dual-source scanner technology

Discussion

Acute PE has a high mortality rate requiring early diagnosis and management.14 In 2014, Al Otair et al1 found a mortality of 8.6% in 105 patients at a University Hospital in Riyadh. They found surgery, obesity and immobility to be the most common risk factors.1 Young and middle age groups (20-60 years) were found mostly affected as indicated by AlGahtani et al2 amongst 341 suspected PE patients. In 2015, Al-Hameed et al3 highlighted recommendations by Saudi Association for Venous Thrombo-Embolism (SAVTE) in assessing clinical probability of DVT before going to sequential diagnostic evaluation. Clinical assessment pathways like Wells score or revised Geneva scores may help to define clinical probability for PE, and needs to be calculated prior to further imaging.15 Results of Prospective Investigation Of Pulmonary Embolism Diagnosis (PIOPED)- II suggested high predictive value of CTPA (utilizing 4-detector CT) in diagnosis of PE in patients with high probability on clinical assessment; however, additional tests might be needed in cases with low clinical suspicion.16 Nevertheless, CTA on fast-speed helical CT has been considered the imaging method of choice for evaluation of PE.17

Improved image quality with reduced dose utilizing dual-source CT scanners (enabling high pitch image acquisition) has been depicted in various studies such as Bauer et al18 in 2012 and Ajlan19 in 2016, and we were able to demonstrate similar results in our study using this scanner technology. However, it should be emphasized that obtaining an optimal image has never been a single-person or a single-step task, rather active involvement of both technologist and radiologist are needed to plan and deliver the right amount of contrast (optimization) at the right time (monitoring) in a right way (delivery technique and system) for the right subject (patient selection), to properly execute image acquisition (scanning parameters and machine settings) and to effectively use post-processing techniques. Patient condition, cooperation and understanding of procedure are also vital to obtain quality imaging. We noted that the number of patients in the older age group (61 and above) were more in dual-source group than in the single-source group (Table 1), these patients are usually unable to hold breath and at times do not follow adequate breathing instructions due to their condition, yet we found a higher sensitivity of pulmonary CTA in Group-II denoting the fact that image quality was not seen hindered by the breathing artifacts. We also observed that CT-true positive cases for PE were seen more in the younger age group (20-40 years) in Group-I (5 patients) versus 3 patients in Group-II, highlighting compliance of breath-hold instructions in these patients facilitating image quality and diagnosis.

Artifacts related to CTPA are well recognized, of these are partial volume artifacts secondary to motion/breathing or cardiac pulsation artifacts, poor contrast opacification due to mixing of opacified and un-opacified blood, beam-hardening artifacts by high density structures (like contrast in SVC and right atrium), confusion with venous structures (venous contamination) and mucous-filled bronchi.20,21 In our study, most of the artifacts in each group were related to breathing that limited the evaluation of subsegmental arteries, and beam-hardening artifacts related to residual concentrated contrast along the SVC limiting evaluation of adjacent upper lobe arteries. We observed that nearly all false positive cases in both groups (3 in group-II versus 5 in group-I) were found to have diagnosis of isolated small subsegmental thrombi, but those patients were seen discharged later on by the clinicians after improvement of symptoms (without anticoagulation) or treatment of cause other than PE, and exclusion of DVT by compression sonography. Although there are debates on clinical relevance of isolated subsegmental thrombosis in stable patients and whether to treat such patients or not, however, presence and diagnosis of such small thrombi may be important and clinically relevant to treat particularly in unstable patients with limited cardiopulmonary reserve.22 Therefore, patient selection (habitus and condition), cannulation access, contrast concentration, flow rate and volume as well as injection protocol technique/optimization (biphasic or triphasic with saline flush) and clear breathing instructions and patient coaching (prior to examination) have been advocated to better image quality and reduction of artifacts.23

It should be noted that radiation dose remains a major concern in utilizing CT studies as these are used in all age groups.24 Therefore, careful review of the clinical information and discussion with referring physician (about clinical assessment scores and laboratory results like D-dimers) are important to ensure proper justification of such studies. Post processing of raw CT data using ‘iterative reconstruction’ has been a topic of interest in the radiology platforms to further minimize dose without compromising image quality.25 We found a 40% reduction in DLPs acquired through the dual source reduced KV and at expense of reducing noise by SAFIRE software. Even further reductions in radiation doses can be possible if dual-energy is also used as studied by Pontana et al26 who documented an even 60% dose reduction using SAFIRE. We emphasize that radiologists should be familiar with the DLP metric information on the scanner console and should monitor this metric in their clinical practice.

Ability of CT to determine embolus burden, location (central or peripheral) and degree of obstruction seem to be important in predicting short-term mortality.4,27 We did not focus on these parameters in our study as these were not our primary objectives. It is also worth noting that certain information like RV function (RV/LV diameter ratio >1.0) and right heart strain (interventricular septal bowing, inferior vena cava reflux, and pulmonary artery diameter exceeding that of aorta) can be easily obtained through these CT studies that could have been added to prognostic value of CT. Computed tomography is also the method of choice for diagnosis and follow up of chronic thromboembolic pulmonary arterial hypertension (CTEPH), demonstrating intravascular web, signs of pulmonary hypertension such as dilatation of pulmonary arteries and RV, mosaic lung parenchymal attenuation, and RV hypertrophy.4 However, as we did not include known or treated patients of pulmonary embolism in our study, therefore, these entities were not discussed in our study. We also excluded pregnant patients with suspicion of PE, because of special consideration (of gestational age, in-utero exposure to contrast and radiation concerns) and possible variability in adaptation of alternative pathway that might be considered in such patients.

Computed tomography pulmonary angiogram has an overall good inter-observer agreement in diagnosis of large and central pulmonary embolism, but there may be significant discordance among radiologists particularly in evaluation of solitary, segmental, and small pulmonary embolism. Discordance of approximately 25% was noted in initial and retrospectively reviewed reports amongst the original radiologist and the chest radiologist by Hutchinson et al.28 Although we did not observe such discordance in our study, yet small defects should always be decided by consensus interpretation to avoid over-diagnosis of small pulmonary embolism. About nearly one third of subsegmental contrast filling defects interpreted by community radiologists were deemed indeterminate and approximately half of these were found to be false positive.29 All pulmonary embolism may not be treated equally like subsegmental pulmonary embolism without DVT in healthy individuals may not be clinically important versus subsegmental emboli with DVT in patients with reserved cardiopulmonary status that should be treated.30 In our study, we did not consider relevance of such diagnosed cases and their prognosis. Also, we did not consider an objective assessment criterion or scoring system for image quality in our study to have more validation of results as were documented in various other studies. Single-center, small sample size and retrospective study are few other limitations of our study.

Certain modifications in CT techniques and scanner technology may be considered in selected patients of chest pain. For exmaple, an ECG-synchronization (as seen in triple-rule-out examinations) and adjustment of contrast material timings may help exclude other causes of acute chest pain (acute aortic syndrome, acute coronary syndrome).4 Utilization of dual-energy CT may help acquire material differentiation and perfusion imaging to increase confidence level for diagnosis of PE.5 Dual-energy perfusion maps display iodine-perfused lung tissue, analogous to pulmonary scintigraphy, and allow assessment of perfusion deficits distal to vessels affected by PE. These may have value in assessing non-obstructive emboli and subsequent risk stratification.31 It would be of interest in future studies to consider utilization of dual-energy CT and acquisition of perfusion imaging and evaluating its impact on diagnosis in both acute and chronic pulmonary cases.

In conclusion, ultra-fast high-pitch dual-source CTPA with free-breathing yields better image quality and diagnosis, as well as reduced image acquisition times and radiation doses when compared to conventional standard-pitch single-source breath-hold CTPA.