The results of this study reveal that the presence of the C.A.P. sign on the CT scan strongly associated with the likelihood of imminent cardiac arrest. The patients with positive C.A.P. sign were more likely to experience cardiac arrest within 1 h after the CT scan. In our study, all the patients with positive C.A.P. sign failed to survive; our result also showed that the C.A.P. sign negatively associated with survival to discharge slightly.
C.A.P. sign occurs owing to the effect of gravity. The density of contrast agent is much higher than that of blood; thus, during normal physiological blood flow, specific gravity has little effect on the contrast agent dynamics. In patients with cardiogenic shock or impending circulatory collapse, both the arterial and venous blood flow dramatically decreases, and the contrast agent tends to accumulate in the dependent parts of the venous system [11]. Additionally, a lack of sufficient blood pressure during shock also inhibits penetration into the organ parenchyma and contributes to the pooling of contrast agent in major vessels [8]. About 52% of the cases demonstrate this effect of gravity after cardiac arrest [12].
In a retrospective study that reviewed 127 patients who underwent contrast-enhanced CT of the chest or abdomen and echocardiography, high injection rate (> 3 ml/sec), tricuspid regurgitation, pulmonary hypertension, and right ventricular systolic dysfunction were found to be an independent predictor of retrograde contrast agent pooling over inferior vena cava or hepatic vein (p< 0.001, < 0.01, = 0.05, and < 0.005, respectively) [13]. This study further demonstrated that the phenomenon of contrast agent pooling is caused by weak venous blood flow and may be influenced by contrast agent injection rate.
We found 25 articles, reporting 59 cases with the sign of dependent contrast agent pooling [2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29]. Age ranges from 6 to 87-year-old. 36 of 59 are male, 21 are female and 2 patients without mentioning of gender. The CT scans with contrast of chest and abdomen were arranged for the survey of trauma (18 cases), aortic dissection, pulmonary embolism, sepsis and abdominal mass, as shown in the supplementary table [Additional file 1].
The site of contrast agent pooling is mostly found over inferior vena cava (IVC). Contrast agent pooling at IVC was described in 55 cases (93.2%). In 4 cases without contrast agent pooling at IVC, 3 cases are traumatic patients with massive hemoperitoneum, one of them with ruptured diaphragm and herniation of liver, the increased intraperitoneal pressure may suppress contrast agent regurgitation in IVC. In second patient without contrast agent pooling was with Type A aortic dissection, contrast agent refluxed at hepatic and lumbar vein. Besides of IVC, contrast agent pooling was also observed in sporadic cases over right hepatic vein, left hepatic vein, dependent liver parenchyma, right renal parenchyma, lumbar vein, splenic vein, superior mesenteric vein, superior vena cava, right atrium, right ventricle, coronary sinus, great cardiac vein, azygos vein and hemiazygos vein.
Thirty-nine of fifty-nine cases (66.1%) were reported to experience cardiac arrest within 30 min after CT. 35 of 39 cases developed cardiac arrest immediately after CT. Among 59 cases with contrast agent pooling sign, there were 10 cases (16.9%) with cardiac tamponade, 5 cases (8.5%) with cardiogenic shock. In 54 cases which mentioned about clinical outcome in literature, 44 cases (81.5%) expired eventually.
There was one case (1.7%) who did not experience cardiogenic shock or cardiac arrest despite of positive contrast agent pooling sign; however, extreme low ejection fraction was noted at follow up echocardiography [20].
Four pediatric cases were found in literature review, three of them were collapsed soon after CT scan [17, 29]. One case with history of chronic constrictive pericarditis developed neither cardiac arrest nor cardiogenic shock was proven with severe right ventricular failure at follow up echocardiography.
Summarized from literature reviews and our cases, the feature of C.A.P. sign includes contrast agent pooling in IVC, right hepatic vein, left hepatic vein, dependent liver parenchyma, right renal parenchyma, lumbar vein, splenic vein, superior mesenteric vein, superior vena cava, right atrium, right ventricle, coronary sinus, great cardiac vein, azygos vein and hemiazygos vein. The most common site of contrast agent pooling is IVC (93.2% from the reported cases, 90.9% from our case series).
In our study, 63.7% of the patients with a positive C.A.P. sign experienced cardiac arrest within 1 h, and all patients with a positive C.A.P. sign experienced cardiac arrest within 8 h. Our result coincides with those of the previous studies, indicating that C.A.P. sign may be used as a predictor for imminent cardiac arrest.
Contrast-enhanced computed tomography may not be a diagnostic tool for impending cardiac arrest and severe cardiogenic shock. Nevertheless, if the patients underwent CT to survey critical illnesses, the physician has an opportunity to inspect for the C.A.P. sign. Recognition of the sign can effortlessly alert primary care physicians, and rapidly indicate the possibility of extremely low cardiac output or impending circulatory failure. Hence, C.A.P. sign should grab physician’s attention to activate timely evaluation and earlier intervention to save the patient’s life. In some cases, earlier termination of CT may be necessary to provide immediate cardiopulmonary resuscitation and prevent end-organ ischemia or death. In clinical scenario, if a patient with positive C.A.P. sign is clinically unstable, we propose earlier intubation and inotropic agent usage to stabilize the patient, and rapidly search and treat potential reversible causes.
This study has several limitations. First, the presentation of dependent venous pooling is not well defined. In some cases, contrast agent pooling was noted to be confined to the right hepatic lobe, right renal vein, inferior or superior vena cava, and even subclavian vein. This condition may imply specific haemodynamic disturbances in these cases.
Second, there are no statistical data on the incidence rate of C.A.P. sign in the normal population or the population with low cardiac output. Further studies are needed to investigate the incidence rate of C.A.P. sign in the normal population and in populations with extremely low cardiac output. Measurement of cardiac output during contrast-enhanced CT may provide more information about the development of the sign.
Third, selection bias. Our study focuses on the CT findings of the patients experienced cardiac arrest in emergency department of single medical centre. The statistic values of C.A.P. sign may not apply to all population. However, as the occurrence of C.A.P. sign in normal population is extremely rare, we believe the accuracy of C.A.P. sign in predicting cardiac arrest would be much higher if the study could include more patients.