| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Pediatric Imaging |
1 Department of Radiology, Division of Pediatric Radiology, Box 3808, Duke University Medical Center, Durham, NC 27710.
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: Sequential isolevel (24-mAs) CT images of the spleen obtained at 6-second intervals after initiation of contrast material injection in 112 children (mean age, 4.5 years) were reviewed. Heterogeneity characteristics assessed included type, onset, maximum, and resolution. Relationship to variables (injection rate, age, splenomegaly) was assessed with the Fisher exact test.
RESULTS: Eighty-one of the 112 patients (72%) had transient heterogeneity: archiform (45 patients), diffuse (25 patients), and focal (11 patients). Mean times were as follows: initial visualization after onset of contrast material injection, 19.2 seconds; maximum heterogeneity, 27.3 seconds; and resolution, 47.4 seconds. Statistically significant relationships were seen between frequency of heterogeneity and injection rate (
1 mL/sec, 82%; <1 mL/sec, 50% [P = .001]), age (>1 year, 76%; 
1 year, 46% [P = .04]), and splenomegaly (present, 20%; absent, 77% [P = .048]).
CONCLUSION: Heterogeneous splenic contrast enhancement is common, has several patterns of appearance, and is predictably encountered during the 70 seconds after the initiation of contrast material injection. Injection rate, age, and presence of splenic disease influence the frequency with which these artifacts are encountered.
Index terms: Children, 775.12115 • Computed tomography (CT), artifact, 775.93 • Computed tomography (CT), helical, 775.12115 • Normal variant, 775.12115 • Spleen, 775.12115 • Spleen, CT, 775.12115
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The rapid speed of helical CT and the current potential to image patients early during the arterial phase of contrast enhancement have led to these artifacts being more commonly encountered. Knowledge of the timing, appearance, and factors affecting the frequency of these heterogeneous enhancement patterns is important for optimizing helical CT protocols and interpreting the importance of low-attenuation lesions seen within the spleen at CT examination.
In children, these issues may be further complicated by the unique and changing composition of the developing spleen (11). We have seen patients for whom imaging has been performed early after the initiation of injection of the contrast material bolus, and resultant heterogeneous splenic enhancement has obscured pathologic lesions demonstrated at repeat spiral CT examination. We have also been referred children with spiral CT images in which heterogeneous splenic enhancement has been interpreted as pathologic lesions; one child underwent splenectomy on the basis of such findings. Because of these cases, we performed the following study.
The purpose of this study was to (a) determine the appearances and timing of heterogeneous splenic enhancement at spiral CT in children and (b) identify influencing variables so that the likelihood of heterogeneity being mistaken for or obscuring splenic pathologic lesions is minimized.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The study group included all patients 20 years of age or younger who underwent CT examination of the abdomen with an intravenous contrast material bolus-tracking protocol (SMARTPREP; GE Medical Systems, Milwaukee, Wis) (12–14) over a 2-year period. The bolus-tracking protocol is used on a clinical basis in our department, in an attempt to maximize the quality of our CT images. One hundred twelve CT scans were included in the study. The mean age of the 112 patients was 4.5 years (range, 2 months to 20 years; 67 male and 45 female patients).
Indications for the CT examinations of patients in the group varied: trauma (n = 16), follow-up of malignant neoplasm (n = 43), ruling out a mass (n = 14), ruling out an abscess in an immunocompetent child (n = 14), ruling out infection in an immunocompromised child (n = 21), and other (n = 4). No indications prohibited inclusion in the study.
All CT examinations were performed with a HiSpeed Advantage system with CT INTUITIVE software (GE Medical Systems, Milwaukee, Wis). All patients received intravenous injection of the nonionic contrast material, iopamidol (Isovue; Bracco Diagnostics, Princeton, NJ) at a dose of 2.0 mL per kilogram of body weight. Contrast material was injected either manually (107 patients) or with a power injector (five patients). In patients for whom the data were available, the rate of contrast material injection (in milliliters per second) was recorded. In patients with manual injection, the rate of injection was calculated by dividing the volume of contrast material administered (in milliliters) by the length of time over which the contrast material was injected (in seconds). For manual injection, the contrast material was injected as fast as possible.
The bolus-tracking images consisted of sequential, isolevel, low-dose CT images obtained at 6-second intervals from the initiation of contrast material injection to the commencement of diagnostic CT imaging. Technical parameters for the bolus-tracking images included 10-mm collimation, 120 kVp, 26 mAs, and soft-tissue algorithm (12–14). In all patients, these images were compared with a noncontrast baseline image obtained at the same level. Region-of-interest cursors were placed over the peripheral portion of the liver and the aorta. Diagnostic imaging was manually initiated when (a) the liver enhanced by 45 HU, (b) liver enhancement reached a plateau, and the aortic enhancement curve was decreasing, or (c) 10 tracking images were obtained (12–14).
The mean number of tracking images was 6.5 (range, 3–10). The radiation dose of a single tracking image was 24% of that of a conventional CT image (12). Therefore, the mean added dose for the 6.5 tracking images is equal to the dose associated with the addition of approximately 1.5 conventional CT sections to a CT examination. The technical parameters for the diagnostic images varied with the size of the patient and included pitch from 1.0 to 1.5, collimation from 5 to 10 mm, 120–140 kVp, and 90–140 mA.
The CT images were evaluated for the presence of transient patterns of heterogeneous enhancement. The images were reviewed simultaneously by two radiologists (J.N.F., L.F.D.), and conclusions were reached by consensus. The reviewers were not blinded to information pertaining to imaging technique or the patient's clinical history.
Splenic heterogeneity was categorized into one of three transient patterns: archiform, focal, or diffuse. Archiform heterogeneity was defined as the presence of alternating bands of high and low attenuation occurring in a ringlike (Fig 1), curvilinear, concentric, zebra-stripe, or bizarre pattern. Focal heterogeneity (Fig 2) was defined as the presence of a focal region of low attenuation compared with the remainder of the enhancing splenic parenchyma. Diffuse heterogeneity was defined by the presence of diffuse, mottled heterogeneous attenuation throughout the spleen without a focal area of low attenuation (Fig 3).
|
|
|
The amounts of time between the initiation of contrast material injection and the initial visualization, maximum visualization, and resolution of splenic heterogeneity (Fig 4) were recorded, and mean times were calculated. The numbers of patients with persistent splenic heterogeneous enhancement persisting at 50, 60, 70, and 80 seconds were also noted.
|
|
|
The relationship of certain variables to the frequency and timing of heterogeneous splenic enhancement was evaluated. Variables evaluated included intravenous contrast material injection rate, age, splenomegaly, and presence of a chronic illness causing prolonged stress. Conditions considered as causes of chronic stress included chronic illness, malignant neoplasm, bone marrow transplant, and other causes of immune suppression. Splenomegaly was defined by a splenic length exceeding the age-based standards described by Rosenberg et al (15).
Subgroups of patients that were compared for correlation with the frequency and timing of heterogeneous splenic enhancement with the Fisher exact test included those defined by contrast material injection rate (1 mL/sec vs <1 mL/sec), age (>1 year vs 1 year), splenomegaly (present vs absent), and chronic stress (present vs absent). Correlations of the frequency of heterogeneity with age and with injection rate were evaluated with a standard Pearson 
2 test. Because young age is often associated with small-bore intravenous access and therefore slow injection rates, a logistic regression model with first-degree freed 2 test was used to evaluate age and intravenous injection rate as independent variables as they relate to the frequency of splenic heterogeneity.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Heterogeneity was characterized by the predominant pattern. The pattern was characterized as predominantly archiform in 45 of the 81 patients with heterogeneity (56%), as diffuse in 25 patients (31%), and as focal low attenuation in 11 patients (14%). The mean time of initial visualization was 19.2 seconds (range, 9–44 seconds). The mean time of maximum visualization was 27.3 seconds (range, 16–53 seconds).
For 13 of the 81 patients with heterogeneity, the diagnostic CT images were obtained and bolus-tracking images stopped before the resolution of splenic heterogeneity. The mean time until the last diagnostic image was obtained through the level of the spleen in these patients was 44.2 seconds (range, 23–57 seconds). These 13 patients were excluded from the calculations for the mean time of resolution of splenic heterogeneous enhancement.
The mean time for resolution of splenic heterogeneous enhancement in the remaining 68 patients was 47.4 seconds (range, 23–86 seconds). Splenic heterogeneity was still persistent at 50 seconds in 31 of these 68 patients in whom resolution of the splenic enhancement was visualized (46%), at 60 seconds in nine patients (13%), at 70 seconds in four patients (6%), and at 80 seconds in one patient (1%).
Because the 13 patients in whom the splenic heterogeneity was not resolved on the final images could be considered censored data when calculating the mean, a Kaplan-Meier estimate was used to predict the distribution for the time to resolution. The estimated mean time of resolution based on this model was 49.0 seconds.
For 77 of the 112 patients, the rate of intravenous contrast material injection was known. For these 77 patients, the mean rate of contrast material injection was 1.60 mL/sec (range, 0.32–4.30 mL/sec). A statistically significant relationship was seen between higher rate of injection and frequency of splenic heterogeneity (P = .001).
For patients with injection rates less than 1.0 mL/sec (n = 22), splenic heterogeneity was identified in 11 (50%), the mean time to initial visualization was 19.2 seconds (range, 16–24 seconds), the mean time of maximum visualization was 26.7 seconds (range, 16–44 seconds), and the mean time until resolution was 47 seconds (range, 28–67 seconds). For patients with injection rates of 1.0 mL/sec or higher (n = 55), splenic heterogeneity was identified in 45 (82%), the mean time to initial visualization was 18.6 seconds (range, 9–31 seconds), the mean time of maximum visualization was 25.6 seconds (range, 16–44 seconds), and the mean time until resolution was 46.0 seconds (range, 23–86 seconds). The difference in the frequency of splenic heterogeneity between these two subgroups of patients was statistically significant (P = .001); however, no statistically significant differences (all P > .1) were seen between the two groups for the mean times of initial visualization, maximum visualization, and resolution of visualized heterogeneity, or for the frequency of patterns visualized.
Statistically significant correlation (P = .05) was seen between increasing age and increased frequency of splenic heterogeneity. For patients older than 1 year of age (n = 99), splenic heterogeneity was present in 75 (76%), the mean time to initial visualization was 19.4 seconds, the mean time of maximum visualization was 27.5 seconds, and the mean time until resolution was 48.3 seconds. For patients 1 year of age or younger (n = 13), splenic heterogeneity was present in six (46%), the mean time to initial visualization was 17.3 seconds, the mean time of maximum visualization was 25.5 seconds, and the mean time until resolution was 40.8 seconds. The difference in the frequency of splenic heterogeneity between these two subgroups of patients was statistically significant (P = .04); however, no statistically significant differences (all P > .1) were seen between the two groups for the mean times of initial visualization, maximum visualization, or resolution of visualized heterogeneity, or for the frequency of patterns visualized.
Injection rates were related to age. The mean rate of intravenous injection of contrast material for patients older than 1 year of age (n = 66) was 1.73 mL/sec (range, 0.45–4.30 mL/sec). The mean intravenous contrast material injection rate for patients 1 year of age or younger (n = 11) was 0.80 mL/sec (range, 0.24–2.00 mL/sec). Because it was suspected that the differences attributed to age may actually only be related to injection rate, age and injection rate were evaluated as independent variables with a logistic regression model. Evidence was found for statistically significant correlation of both increasing injection rate (P = .001) and age (P = .007) independently in relationship to the frequency of splenic heterogeneity.
Splenic heterogeneity was seen in two of 10 patients (20%) with splenomegaly, compared with 79 of 102 patients (77%) without splenomegaly. The difference between the two groups was statistically significant (P = .048). Splenic heterogeneity was seen in 46 of 64 patients (72%) who were considered chronically stressed, compared with 35 of 48 patients (73%) who were not. The difference between the two groups was not statistically significant (P > .1). In addition to the 10 patients who had splenomegaly, three patients had small splenic lacerations. In all three patients, the lacerations were not in the same plane in which the tracking images were obtained. No other splenic abnormalities were present.
The relationships between the frequency of heterogeneity and the subgroups of patients are summarized in the Table.
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Splenic heterogeneity should be anticipated when helical CT examinations are performed during the 1st minute after the commencement of contrast material injection. Patterns of splenic heterogeneity included archiform (alternating bands of low and high attenuation), focal, or diffuse heterogeneity.
In this study, transient splenic heterogeneity was very common, occurring during bolus tracking in 72% of the examinations. The prominence of this heterogeneity was greater in the early period after the initiation of contrast material injection. The mean time of maximal visualization was 27 seconds after the initiation of contrast material injection. In over 95% of the patients, the splenic heterogeneity resolved by 70 seconds. Therefore, low-attenuation "lesions" seen within the spleen on spiral CT images obtained 70 seconds after the commencement of contrast material injection should be considered much more likely to be pathologic lesions than low-attenuation areas of the spleen seen on images obtained during the first 60 seconds after contrast material injection.
The frequency of transient splenic heterogeneity is influenced by several factors. In this study, a statistically significant relationship was seen between increased rate of contrast material injection and the frequency of transient splenic heterogeneity (P = .001). Helical CT examinations performed with a contrast material injection rate of 1.0 mL/sec or higher had a much greater frequency of splenic artifacts than did those with a contrast material injection rate of less than 1.0 mL/sec. In children, contrast material injection rates can vary more than in adults because intravenous access of various sizes is used and because hand injection is often used. The small caliber of the intravenous cannula limits the rate of injection. In this study, injection rates varied from 0.32 to 4.30 mL/sec. Therefore, a low-attenuation area seen within the spleen of an infant whose CT was performed after manual injection of contrast material through small-caliber intravenous access at a slow rate of injection must be viewed with greater suspicion than a low-attenuation area seen on a CT image obtained after power injection of contrast material at a rapid rate. Of importance, however, the rate of contrast material injection did not affect the time of visualization, the time of resolution, or the pattern of splenic heterogeneity. These parameters are not dependent on rate.
Because splenic enhancement patterns are related to the unique anatomy of the spleen and because the anatomy of the developing spleen changes with age, we suspected that the frequency of splenic enhancement patterns might also be related to the age of the patient. The differential flow rates of blood to the lymphoid follicles (white pulp) and vascular sinusoids (red pulp) are thought to create these splenic enhancement patterns (1–3,7). The ratio of white pulp to red pulp varies with age, with a paucity of white pulp in infants (11). During the 1st year of life, the ratio of white pulp to red pulp gradually increases. These changes in the histologic appearance of the spleen have been shown to affect its signal intensity on magnetic resonance images (11). In this study, a statistically significant relationship was seen between increased age and increased frequency of splenic heterogeneity (P = .04). This difference in frequency by age may be related to a requirement for adequate amounts of white pulp to generate over time to produce differential flow patterns within the spleen. Although the mechanism is speculative, it is clear that heterogeneous splenic enhancement is less common in young children.
In adults, several other factors affect the presence of heterogeneous splenic enhancement, such as the presence of liver disease, portal vein thrombosis, and congestive heart failure (2,7). None of the patients in this study had these problems; however, we wondered if other factors that might affect the splenic composition (white pulp–to–red pulp ratio) might also affect the frequency of splenic heterogeneity. A decreased frequency of heterogeneity was seen in children with splenomegaly (P = .048); however, chronic stress, which we suspected would deplete white pulp reserves and result in a lack of heterogeneous splenic enhancement, was not associated with a statistically significant trend (P > .1).
This study had several limitations. First, all subjects in this study were referred for helical CT examination because of a known or suspected abnormality. Therefore, the study population in which the timing and appearance of splenic heterogeneity was defined does not represent healthy children. We believe that this is only a minor limitation both because it is unlikely that CT examinations of the healthy pediatric population could be justified and because our study population reflects the kind of child in whom CT examinations are usually performed. In fact, this population is the one for which this information is most pertinent. The second potential limitation is that the sequential isolevel images used in this study were obtained as part of a bolus-tracking technique, tailored to determining the appropriate timing of CT imaging of the liver. The images were not obtained in a fashion tailored to evaluate the temporal course of splenic enhancement. In 13 of the 112 patients, diagnostic imaging was commenced and bolus-tracking images stopped before complete resolution of the splenic heterogeneity. In addition, in patients who underwent manual injection, the duration of the injection was not available in all cases. Finally, because evaluation of the pathologic lesions of the spleen was not available for the patients in this study, the standard of reference for proving that this heterogeneity truly represented flow artifacts was based on the temporal course, resolution, and pattern of the heterogeneity, as well as the absence of focal splenic lesions seen on diagnostic examinations.
In conclusion, transient splenic heterogeneity is a normal flow phenomenon, related to the unique architecture of the spleen, and is commonly encountered during the first 70 seconds after the initiation of contrast material injection. Factors that are associated with an increased frequency of these flow artifacts in children include rapid injection rates, increasing age, and absence of splenomegaly. These same factors, however, do not affect timing of onset and resolution of heterogeneity. Low-attenuation lesions encountered in the spleen on CT images obtained within 70 seconds after initiation of contrast material injection, when the injection rate is slow, in children younger than 1 year of age, or in the presence of splenomegaly must be viewed with a high index of suspicion for pathologic lesions.
|
Footnotes |
|---|
Author contributions: Guarantors of integrity of entire study, L.F.D., J.N.F., D.P.F.; study concepts, J.N.F., L.F.D., D.P.F., G.S.B.; study design, L.F.D.; definition of intellectual content, L.F.D., J.N.F., D.P.F., G.S.B.; literature research, L.F.D.; clinical studies, L.F.D., J.N.F., D.P.F., G.S.B.; data acquisition, J.N.F., L.F.D., D.P.F.; data analysis, J.N.F., L.F.D.; statistical analysis, J.N.F., L.F.D.; manuscript preparation, L.F.D.; manuscript editing and review, L.F.D., J.N.F., D.P.F., G.S.B.
Received May 27, 1998;
revision requested June 24, 1998; revision received July 22, 1998;
accepted September 28, 1998.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Paterson, D. P. Frush, L. F. Donnelly, J. N. Foss, S. M. O'Hara, and G. S. Bisset III A Pattern-oriented Approach to Splenic Imaging in Infants and Children RadioGraphics, November 1, 1999; 19(6): 1465 - 1485. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
