After myocardial ischemia and stroke, pulmonary embolism (PE) is the most common cause of cardiovascular death and represents a major burden in healthcare.1 Computed tomography (CT) pulmonary angiography (CTPA) is currently the criterion standard for diagnosing PE.2
Although CTPA remains the current criterion standard for diagnosing PE, technological advances in multidetector row computed tomography (MDCT)—higher acquisition speed, thinner collimations, increased spatial resolution, and virtual monoenergetic images—have led to increased findings of incidental PE on routine contrast-enhanced CTs.3,4 In one study, 54% of all PEs in cancer patients and 19% of all PEs in noncancer patients were diagnosed incidentally on CT examinations performed for reasons outside of clinically suspected PE.5 Another study reported that 54% of diagnosed PEs were incidental findings on routine contrast enhanced CTs.6 In a retrospective review, Storto et al3 found that a significant proportion of routine contrast-enhanced MDCTs (3.4%) demonstrated unsuspected PE. den Exter et al7 reported that unsuspected PE diagnosis on routine chest CT was reliable and reproducible despite the suboptimal reading conditions of a nondedicated scan protocol. These findings suggest a possible role for routine chest CTs in the diagnosis of incidental PE.
Despite advances in MDCT technology and increased findings of incidental PE on routine chest CT, questions remain about the reliability of these scans for incidental PE diagnosis. In a prospective study of oncology patients undergoing both routine and PE-protocoled CT, Browne et al8 found that 39% of unsuspected PE cases could only be diagnosed with CTPA. Similarly, Gladish et al9 reported that only 25% of unsuspected PE was reported at the initial image interpretation. Pitfalls to diagnosis include the use of inappropriately narrow window settings, thicker image slices, and the failure of the radiologist to be cognizant of PE when reading routine contrast-enhanced CT.10
Although routine contrast-enhanced chest CTs appear to have a role in diagnosing incidental PE, further investigation is necessary. The purpose of the present study was 2-fold: first, to determine the average attenuation of the pulmonary arteries in a year's worth of routine contrast-enhanced chest CTs at a multisite urban academic center, and second, to determine the percentage of those routine chest CTs that meet the threshold of 200 Hounsfield units (HU) attenuation in the pulmonary arteries, which most radiologists would consider adequate enhancement to evaluate for PE.
MATERIALS AND METHODS
All routine contrast-enhanced chest CTs performed on adults 18 years or older from a multisite urban academic medical center from January 2018 to December 2018 were retrospectively identified. Computed tomographies were excluded if they were performed as part of a CT angiogram protocol for suspected PE or for other indications like aorta studies; gastrointestinal bleed studies; liver, pancreas, and renal CT angiogram studies; and electrocardiographically gated CTs. For patients who were scanned more than once during the study period, each scan was included.
Computed tomographies were performed at 4 emergency departments, 3 inpatient hospitals, and several outpatient sites, most often as part of a routine chest, abdomen, and pelvis CT examination.
The routine chest CTs were performed on a variety of scanners—most commonly GE Lightspeed 64-detector row CT (GE Healthcare, Chicago, IL) with administration of 100 to 140 mL of iopamidol 61% at 1 to 3 mL/s depending on the scan range, indication, and caliber of the intravenous line. No saline chaser was administered. Scans were obtained in craniocaudad direction.
All CTs had axial 5-mm thick sections and sagittal and coronal 3-mm-thick sections routinely archived on Picture Archiving and Communication System in standard kernel reconstruction. Pulmonary artery enhancement was measured on axial images by placement of regions of interest in the main, right and left main, right interlobar, and left lower lobe pulmonary arteries (Figs. 1, 2). For the purposes of this study, we divided the central pulmonary arteries into the main pulmonary artery and the more peripheral, but still central, pulmonary arteries: the right and left main pulmonary artery, the right interlobar pulmonary artery, and the left lower lobar pulmonary artery. We wanted to evaluate in this study whether there was any difference in enhancement between the main pulmonary artery and the more peripheral, but still central, pulmonary arteries. Adequate enhancement was defined as a threshold of 200 HU or greater average enhancement within the main pulmonary artery. The threshold of 200 HU was selected based on prior studies in the literature.11–13
Electronic health records were reviewed for patient age, sex, location, indication for imaging, and administered contrast volume. Computed tomography reports were reviewed for the diagnosis of PE. We did not reevaluate the images of CTs reported as positive for PE. The objective of the current study was not to compare the accuracy of nonthoracic subspecialty trained radiologists for the CT diagnosis of PE with those who have subspecialty training in thoracic radiology.
Descriptive statistics were calculated for enhancement in both the main pulmonary artery and for the aggregate of peripheral pulmonary arteries. Pearson correlation coefficients were used to evaluate the correlation between the attenuation in the main pulmonary artery and attenuation of the peripheral pulmonary arteries. The association between patient factors and CT examinations meeting the 200 HU threshold was evaluated with multiple logistic regression analysis. We assessed interactions among the 4 factors in the logistic regression and found no statistical evidence of interactions. A χ2 test was performed to evaluate the association between PE-positive diagnosis and the 200 HU threshold.
The institutional review board approved this study. Informed consent was not required.
A total of 3164 CT scans met the inclusion criteria. The mean patient age was 63.2 years (SD, 14.2; range, 19–104 years), and 55.8% (1764 of 3164) were women. The 200 HU threshold was met in 28.7% (907 of 3164) of studies. The main pulmonary artery mean attenuation was 191.6 HU (SD, 47.9 HU; range, 51.9–785.6 HU), and the peripheral pulmonary artery mean attenuation was 181.2 HU (SD, 45.9 HU; range, 40.4–694.4 HU) (Table 1). Attenuation within the main pulmonary artery was highly correlated with the mean peripheral pulmonary artery attenuation (r = 0.965) (Fig. 3).
TABLE 1 -
Pulmonary Vasculature Enhancement Within the Main Pulmonary Artery (Main PA) and the Peripheral Pulmonary Artery (Peripheral PA)
|Pulmonary Arterial Vasculature
|Main PA (all cases)
|Peripheral PA (all cases)
|Pulmonary Arterial Vasculature
|Main PA (PE+ cases)
|Peripheral PA (PE+ cases)
The first table shows data for all patients regardless of PE status. The second table shows data for PE-positive cases only.
PA indicates pulmonary artery.
Four patient factors were associated with CT scans meeting the 200 HU threshold: female sex (odds ratio [OR], −1.86); older age, per 10-year increment (OR, −1.28); outpatient status versus emergency department status (OR, −1.52), outpatient status versus inpatient status (OR, −1.57); and volume of administered contrast per 10 mL increment (OR, −1.24).
Pulmonary embolism was diagnosed in 1.8% (58 of 3164) of routine chest CT scans. Furthermore, 39.7% (23 of 58) reached the 200 HU threshold. For PE-positive cases, the main pulmonary artery mean attenuation was 202.4 HU (SD, 45.3 HU; range, 133.2–330), and the peripheral pulmonary artery mean attenuation was 193.1 HU (SD, 44.9 HU; range, 127.7–332.6 HU) (Table 1).
There was a trend toward a higher proportion of studies meeting the 200 HU threshold for patients diagnosed with PE compared with patients without PE (39.7% [23 of 58] vs 28.5% [884 of 3106], P = 0.085) (Fig. 3).
The present study of 3164 routine contrast-enhanced CT scans, to our knowledge the largest series to date, evaluated the enhancement of the pulmonary arteries on routine contrast-enhanced chest CTs. Furthermore, 28.7% of all studies and 39.7% of PE-positive cases reached the 200 HU threshold indicative of adequate enhancement. This information would be helpful for a clinician taking care of a patient with a new concern for PE. If the diagnosis of PE is newly being considered by the clinician and the patient had a recently performed routine contrast-enhanced chest CT, ordering a subsequent CT pulmonary angiogram may not be necessary. Instead, careful review of recently performed routine contrast-enhanced CT chest may be adequate. This would reduce the burden of radiation, iodinated contrast exposure, cost, and time spent in hospital from additional testing. A workflow that prioritizes careful image review of a recently performed examination (often as a chest, abdomen, pelvis CT), rather than reflexively performing a CTPA, is warranted when a question of PE arises. Enhancement of the pulmonary arterial vasculature is adequate for diagnosis in a substantial minority of cases.
Several patient factors were correlated with studies meeting the 200 HU threshold: female sex, older age, outpatient status, and increased contrast amount administered. We posit that female sex is correlated with adequate attenuation because women generally have a smaller body habitus and lower blood volumes compared with men, leading to higher contrast enhancement.14 Older patients generally have a less dynamic cardiovascular system leading to less contrast washout and easier timing of contrast administration. The lower cardiovascular output states of older patients lead to prolonged peak enhancement for these patients.14 Outpatients tend to be able to more easily cooperate with examinations compared with ED patients or inpatients. Finally, increased volume of administered contrast generally leads to greater enhancement within the pulmonary arterial vasculature. Additional factors besides pulmonary artery enhancement, such as the patient's ability to breath-hold, artifacts, and presence or absence of adjacent lung abnormality can also impact PE diagnosis but were not the focus of the present study.15 Patient breathing, for instance, can lead to respiratory artifacts, and thus, patients should perform an inspiratory breath-hold. Patients should also be advised not to perform a Valsalva maneuver, which can lead to bolus interruption and lower vascular enhancement due to an influx of unopacified blood from the inferior vena cava.16
Interestingly, for cases where a PE was diagnosed, the median attenuation was 191.5 HU for the main pulmonary artery (range, 133.2–330 HU) and 185.5 HU for the peripheral pulmonary arteries (range, 127.7–332.6 HU) (Fig. 4). Therefore, in more than half of the PE-positive cases, the radiologist diagnosed PE despite the attenuation being below the 200 HU threshold. This suggests that the 200 HU threshold for attenuation within the pulmonary vasculature is not absolutely necessary. Figure 5 shows an example of a PE found in a routine chest CT with suboptimal attenuation in the pulmonary vasculature (Fig. 5). Some authors have used 180 HU or even 100 to 150 HU as the lower limit of acceptable attenuation for evaluation of PE in the majority of the population.3,17–19 An additional image shows an example of a PE found in a routine chest CT with optimal attenuation in the pulmonary vasculature—in this case, the attenuation was 267 HU within the main pulmonary artery (Fig. 6).
The use of routine chest CT to diagnose unsuspected PE can have therapeutic consequences. The accurate diagnosis of PE, including incidental PE, can lead to proper treatment and management of the patient. Although the benefit of treating unsuspected PE is still unclear, the majority of clinicians opt to begin therapy in patients where the benefits outweigh the risks.20 In addition, unsuspected PE is associated with adverse survival and increased long-term mortality, leading to the general consensus to start anticoagulation therapy in cancer patients with unsuspected PE.20
This study had several limitations. It is not possible to know how many PEs were present but not diagnosed; the diagnosis of PE was based on the radiology report. In addition, thin section images were not routinely reviewed or archived on PACS, and hence, they were not available for this retrospective study; however, in current clinical practice, thin-section axial images are often available for review, either archived on PACS or on a workstation. Body surface area, body mass index, contrast injection rate, the gauge or location of the IV insertion site, and the type of CT scanner used were not accounted for, which are all factors that can impact image quality.21 We also did not evaluate the contrast-to-noise ratio or signal-to-noise ratio. Future studies should incorporate other patient populations and account for these additional factors that can impact study quality. Another potential limitation is that only the main pulmonary artery and the more proximal peripheral arteries were examined. However, although this study only measured the more proximal pulmonary arteries, literature suggests that it is unnecessary to treat isolated subsegmental PE in stable patients with low risk of recurrent PE.22
In conclusion, over one quarter of the 3164 evaluated routine contrast-enhanced chest CT scans met the 200 HU threshold indicative of adequate pulmonary artery enhancement for evaluation of PE. If a routine contrast-enhanced chest CT has recently been performed and concern for PE subsequently arises, directed rereview of the routine CT may be adequate to answer the clinical question, thus avoiding the necessity for repeat imaging with dedicated CTPA.
1. Wu CC, Gilman MD. Chapter 50: Acute pulmonary embolism. In: Walker CM, Chung JH, eds. Muller's Imaging of the Chest
. 2nd ed.Philadelphia, PA: Elsevier, 2019: 622–632.
2. Stein PD, Fowler SE, Goodman LR, et al. Multidetector computed tomography for acute pulmonary embolism. N Engl J Med
3. Storto ML, Di Credico A, Guido F, et al. Incidental detection of pulmonary emboli on routine MDCT of the chest. AJR Am J Roentgenol
4. Bae K, Jeon KN, Cho SB, et al. Improved opacification of a suboptimally enhanced pulmonary artery in chest CT: experience using a dual-layer detector spectral CT. AJR Am J Roentgenol
5. Giuntini C, Di Ricco G, Marini C, et al. Pulmonary embolism: epidemiology. Chest
. 1995;107(1 Suppl):3S–9S.
6. Bach AG, Meyer HJ, Taute BM, et al. The frequency of incidental pulmonary embolism in different CT examinations. Br J Radiol
7. den Exter PL, van der Hulle T, Hartmann IJ, et al. Reliability of diagnosing incidental pulmonary embolism in cancer patients. Thromb Res
. 2015 Sep;136:531–534.
8. Browne AM, Cronin CG, English C, et al. Unsuspected pulmonary emboli in oncology patients undergoing routine computed tomography imaging. J Thorac Oncol
. 2010 Jun;5:798–803.
9. Gladish GW, Choe DH, Marom EM, et al. Incidental pulmonary emboli in oncology patients: prevalence, CT evaluation, and natural history. Radiology
. 2006 Jul;240:246–255.
10. Tresoldi S, Flor N, Luciani A, et al. Contrast enhanced chest-MDCT in oncologic patients. Prospective evaluation of the prevalence of incidental pulmonary embolism and added value of thin reconstructions. Eur Radiol
. 2015 Nov;25:3200–3206.
11. Hendriks BMF, Eijsvoogel NG, Kok M, et al. Optimizing pulmonary embolism computed tomography in the age of individualized medicine. Invest Radiol
12. Gossner J. Feasibility of computed tomography pulmonary angiography with low flow rates. J Clin Imaging Sci
13. Bae KT. Optimization of contrast enhancement in thoracic MDCT. Radiol Clin North Am
14. Digumarthy SR, Padole A, Tabari A, et al. Pulmonary CT: the scanner, the protocol, and the dose. In: Digumarthy SR, Abbara S, Chung JH, eds. Problem Solving in Chest Imaging
. Philadelphia, PA: Elsevier; 2020:76–90.
15. Wittram C. How I do it: CT pulmonary angiography. AJR Am J Roentgenol
16. Albrecht MH, Bickford MW, Nance JW Jr., et al. State-of-the-art pulmonary CT angiography for acute pulmonary embolism. AJR Am J Roentgenol
17. Ozawa Y, Hara M, Shibamoto Y. The frequency of insufficient contrast enhancement of the pulmonary artery in routine contrast-enhanced chest CT and its improvement with an increased injection rate: a prospective study. J Thorac Imaging
18. Delesalle MA, Pontana F, Duhamel A, et al. Spectral optimization of chest CT angiography with reduced iodine load: experience in 80 patients evaluated with dual-source, dual-energy CT. Radiology
19. Engelke C, Manstein P, Rummeny EJ, et al. Suspected and incidental pulmonary embolism on multidetector-row CT: analysis of technical and morphological factors influencing the diagnosis in a cross-sectional cancer centre patient cohort. Clin Radiol
20. den Exter PL, Jiménez D, Kroft LJ, et al. Outcome of incidentally diagnosed pulmonary embolism in patients with malignancy. Curr Opin Pulm Med
21. Kalra MK, Otrakji A, Digumarthy SR. Radiologic techniques and dose reduction strategies in chest imaging. In: Shepard JO, ed. Thoracic Imaging the Requisites
. 3rd ed. Philadelphia, PA: Elsevier; 2019:1–18.
22. Chiu V, O'Connell C. Management of the incidental pulmonary embolism. AJR Am J Roentgenol