| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Emergency Radiology |
1 From the Departments of Radiology (A.I.R., P.J.K., M.E.L., D.E.M., J.K.S.), Epidemiology and International Health (G.M.), and Surgery-General Surgery (S.T.W.), University of Alabama at Birmingham, 619 S 19th St, Birmingham, AL 35249. Received October 16, 2002; revision requested December 23; final revision received August 4, 2003; accepted August 6. Address correspondence to P.J.K. (e-mail: pkenney@uabmc.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONAPPENDIX: Injury Severity ScoreREFERENCES |
|---|
MATERIALS AND METHODS: The radiology information system and the trauma registry were searched for cases of adrenal hematoma detected at trauma CT during a 54-month period. CT images depicting adrenal masses with the published characteristics of adrenal hematoma were reviewed by readers who were unblinded to the initial interpretations. Injury severity score (ISS), associated injury, and patient outcome data were gleaned from the trauma registry. The control group comprised patients entered in the trauma registry during the study period who did not have a diagnosis of adrenal hematoma.
RESULTS: Fifty-four adrenal hematomas were detected in 51 patients: 42 with right-gland, 12 with left-gland, and three with bilateral lesions. The hematomas were round or ovoid and had a mean maximum diameter of 2.8 cm ± 0.7 (SD) and a mean attenuation of 52 HU ± 12. Periadrenal stranding was seen with 48 (89%) hematomas. At follow-up CT, 32 of 35 hematomas had resolved or decreased in size and attenuation. One patient with adrenal hematoma had no other intraabdominal injuries. Compared with the 6,757 control patients, the 51 patients with adrenal hematoma had a higher mortality rate (10% vs 4%; P < .001, 
2 test) and a higher mean ISS (25.2 vs 9.7; P < .01, t test). Adrenal hematoma was found in 24 (0.4%) of 5,665 trauma patients with an ISS of 0–19, as compared with six (5.0%) of 122 patients with an ISS of 40 or higher.
CONCLUSION: Adrenal hematoma was detected in 51 (1.9%) of 2,692 trauma patients who underwent CT, or 0.8% of all patients (n = 6,808) entered in the trauma registry. Compared with the other trauma patients, the patients with adrenal hematomas had severe injuries associated with higher mortality.
© RSNA, 2004
Index terms: Adrenal gland, CT, 86.12111, 86.12112, 86.12114, 86.12115 • Adrenal gland, hemorrhage, 86.413 • Emergency radiology • Trauma
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONAPPENDIX: Injury Severity ScoreREFERENCES |
|---|
The use of CT in the trauma setting has changed substantially since the report of Burks et al in 1992 (7). Owing to recognition of the diagnostic accuracy of CT, a CT scanner was installed directly in the emergency department at our institution, like at many trauma centers. This unit enables CT evaluation of even severely injured patients. On the other hand, the clinical threshold for requesting CT evaluation has lowered to the extent that patients with even minimal injuries may undergo CT before being discharged rather than be admitted for observation. Our experience has been that adrenal hematomas are not infrequently encountered.
The purpose of our study was to evaluate the frequency of the detection of trauma-induced adrenal hematomas in current practice by using CT and to correlate the presence of adrenal hematoma with the quantitative clinical indicators of injury severity.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONAPPENDIX: Injury Severity ScoreREFERENCES |
|---|
For a control group, we chose all patients entered in the trauma registry during the study period who did not have a diagnosis of adrenal hematoma. During this 54-month period, a total of 6,808 patients were entered in the trauma registry, and 2,692 of them underwent CT of the abdomen. Thus, our study population included 51 patients with adrenal hematoma and 6,757 control patients (2,641 of whom underwent abdominal CT).
All CT examinations were performed by using a GE LightSpeed or GE HiSpeed scanner (GE Medical Systems, Milwaukee, Wis), usually after oral and intravenous contrast material administration. Each patient received 150 mL of nonionic contrast material intravenously at a rate of 2–3 mL/sec by means of a power injector and 12 ounces of oral contrast material. The examinations were performed in a helical mode with 5-mm collimation and 11.25-mm table speed. Depending on the clinical concern(s), the patients underwent CT of the chest, abdomen, and pelvis, and sometimes CT of the abdomen and pelvis only. Patients who were clinically suspected of having head injury also underwent head CT.
The usual image acquisition times after contrast material injection were 45 seconds for the chest and 75 seconds for the abdomen, with a 180-second pause at the lower part of the abdomen before pelvic scanning. If renal injury was suspected, additional images of the abdomen were obtained after a 3–5-minute delay. With the exception of CT performed in two patients who did not receive intravenous contrast material for unrecorded reasons (but presumably because of a history of adverse reaction or renal insufficiency), all examinations were performed in this manner.
CT Reports and Image Review
The reports and film hard-copy images from all CT examinations with findings positive for adrenal hematoma were reviewed by two readers (A.R., P.J.K.) together during the same session. These readers were unblinded as to the initial image interpretations, and all decisions were made by consensus. The available soft-copy images were reviewed at a workstation. Cases in which the initial CT reading indicated adrenal hematoma were considered positive for the disease only when an adrenal mass lesion with characteristics consistent with the published characteristics of adrenal hematoma—specifically, a round or oval high-attenuation mass either occupying the adrenal gland region or obscuring the normal adrenal gland, with or without periadrenal stranding (7)—was identified at retrospective review.
Cases in which there was a normal-appearing adrenal gland but a hematoma in the adrenal gland region that could be explained by hepatic or renal injury were excluded. Cases in which the initial CT report described an adrenal mass but not clearly a hematoma were reviewed, and if the image findings and follow-up data indicated an alternative diagnosis such as adenoma, the case was excluded. Cases in which the initial CT report described the adrenal glands as normal were also excluded, and the images in these cases were not reviewed.
For all patients with CT findings positive for adrenal hematoma, data were collected and recorded on a study form. These data included the size of the adrenal hematoma measured on either film hard-copy or soft-copy CT images and, for patients with available soft-copy images, the attenuation measurements obtained by using a region-of-interest cursor encompassing as much of the hematoma as possible on the section on which the lesion appeared to be the largest. The appearance of the surrounding structures—for example, strandlike attenuation similar to soft tissue in the adjacent fat or active extravasation of intravascular contrast material—and any associated abdominal, chest, head, or orthopedic injuries were also recorded.
For the patients with adrenal hematoma, all of whom underwent follow-up CT (because this was a retrospective study, all follow-up images were obtained for clinical indications), the CT images were reviewed by the same two readers (A.R., P.J.K.) in the same unblinded consensus manner. The sizes, shapes, and changes in attenuation of the adrenal hematomas, as well as other associated or new findings on these follow-up images, were recorded. Trauma registry data on every patient who underwent CT were available and were used to record mechanism of injury, injury severity score (ISS) (8), length of stay in the intensive care unit and/or hospital, and outcome (ie, survival or death). We chose ISS and outcome as the primary objective clinical indicators of injury severity because they are widely used and reproducible measures (see Appendix).
Statistical Analyses
The difference in mortality rate between the study group (ie, 51 patients in trauma registry with diagnosis of adrenal hematoma) and the control group (ie, 6,757 patients in trauma registry without diagnosis of adrenal hematoma) was evaluated by using the 2 test. The mean ISS and the mean lengths of stay in the hospital and intensive care unit for the study and control groups were calculated and compared by using the t test. In addition, an analysis of the study and control groups stratified by ISS was also performed to further investigate the relationship between injury severity and finding of adrenal hematoma. P values of less than .05 were considered to indicate statistically significant differences. In all analyses, statistical evaluations were performed on a patient rather than hematoma basis—that is, for the purposes of this study, a patient was considered to have disease if he or she was found to have one or more adrenal hematomas and not to have disease if he or she was found to have no adrenal hematomas. Thus, multiplicity did not affect the statistical evaluation.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONAPPENDIX: Injury Severity ScoreREFERENCES |
|---|
Forty-two patients had right-sided (Fig 1) and 12 had left-sided (Fig 2) adrenal hematomas. Bilateral adrenal hematomas were identified in three patients (Fig 3). The age range of these patients was 17–75 years (mean age, 40.6 years ± 17.0 [SD]). Their mean age was not statistically different from that of the control group (mean age 36 years ± 16.4). There were 36 (71%) male and 15 (29%) female patients in the study group. Seventy percent (4,704/6,737, 20 age data points were missing) of the control patients were male. The mechanisms of injury for the study group were as follows: motor vehicle collision in 39, fall in two, gunshot wound in one, explosion in one, industrial accident in one, and unspecified in seven patients.
|
|
|
|
|
|
|
|
|
All of the patients with adrenal hematoma were hospitalized. Their lengths of stay in the hospital ranged from 1 to 370 days (mean, 24.79 days ± 53.8). Thirty-six of the 51 patients with adrenal hematoma spent between 1 and 48 days (mean, 10.6 days ± 13.4) in the intensive care unit. Many of the 6,757 patients without adrenal hematoma were not hospitalized. For the control group, the mean length of stay in the hospital was 6.51 days ± 16.10 (range, 0–380 days) and the mean length of stay in the intensive care unit was 2.33 days ± 8.37 (range, 0–139 days). Differences in mean lengths of stay in the hospital and intensive care unit between the study and control groups were significant (P < .001).
There were five (10%) deaths among the 51 patients. The mortality rate in the control group was 4% (253 of 6,757 patients). The difference in mortality rate between the study and control groups was statistically significant (P < .001). The mean ISS for the patients with adrenal hematoma was 25.2 ± 13. The 6,757 patients in the trauma registry without adrenal hematoma had a mean ISS of 9.7 ± 9 (P < .01). In the ISS-stratified analysis, the frequency with which adrenal hematoma was found increased with higher ISS category (Table 2). Of the control patients, 83.5% had an ISS lower than 20, and 1.7% had an ISS of 40 or higher. Of the 51 patients with adrenal hematomas, 47% had an ISS lower than 20, while 12% had an ISS of 40 or higher. Of all the trauma patients seen during the study period with an ISS of 0–19 (n = 5,665), 0.4% had adrenal hematoma; of those with an ISS of 20–39 (n = 765), 2.6% had adrenal hematoma; and of those with an ISS of 40 or higher, 4.9% had adrenal hematoma.
|
We searched the University of Alabama at Birmingham Hospital medical records for cases of potential adrenal insufficiency, but long-term information was available for less than one-third of the patients (many individuals had undergone follow-up at another facility or were lost to follow-up), and in those with available long-term information, no testing for adrenal insufficiency was performed.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONAPPENDIX: Injury Severity ScoreREFERENCES |
|---|
Thus, many institutions, including ours, have taken the opportunity when redesigning or expanding their trauma units to install a CT scanner that is dedicated for use in evaluating trauma patients and other patients treated in the emergency department, with the scanning room equipped for resuscitation, if necessary. This has enabled the use of CT for the evaluation of severely injured patients without compromises to their care.
In our study, we identified 54 adrenal hematomas, in 51 trauma patients, that occurred over a period of 54 months. However, over the duration of the study, a total of 6,808 patients, 2,692 of whom underwent trauma CT, were seen by trauma service personnel. Thus, the frequency of the detection of adrenal hematoma in patients who underwent trauma CT in this study, 51 (1.9%) of 2,692 patients, is surprisingly similar to the frequency of 2% reported in the Burks et al study (7), despite the availability of CT for the evaluation of even severe trauma patients. When the denominator was all patients who were examined for trauma, the frequency of adrenal hematoma detection was only 0.8% (51 of 6,808 patients) in our series. One possible explanation for this seeming paradox could be that there has been, concomitant with the increased use of CT to evaluate patients with more severe injuries, an increased tendency to perform CT to evaluate patients with less severe injuries compared with the frequency of performing CT in these patients in the past.
Burks et al (7) did not report the indications for CT or describe in detail the characteristics of the control group (which also consisted of trauma patients who were examined during the study period but in whom adrenal hematoma was not detected at CT) in their study. The trauma physicians at our institution have developed a very low threshold for requesting CT to examine trauma patients. In the present study, although not all of the trauma patients underwent abdominal CT, almost all of those with substantial injuries did. Even the patients who had immediately undergone surgery underwent follow-up CT when they were stable. Given these factors, it seems likely that few, if any, adrenal hematomas were missed in those patients in whom CT was not warranted on clinical grounds.
Sevitt (1) noted from an autopsy series that adrenal hematomas tended to be found in patients who had severe injuries, and, thus, those with adrenal hematoma had a higher incidence of other injuries, including hepatic, splenic, renal, chest, orthopedic, and central nervous system injuries. Our study results show that there was a significant difference in ISS between the patients who had adrenal injuries (25.2) and those who did not (9.7) (P < .01). The analysis results stratified by ISS are revealing: Less than 17% of patients in the control group had an ISS higher than 20, but at least 50% of those in the study group (ie, with adrenal hematoma) had an ISS higher than 20. At least one adrenal hematoma was found in 5% of patients with an ISS higher than 40.
There was also a significantly higher mortality rate among the patients with adrenal injuries (10% vs 4% in the control group, P < .001). In our series, 26 patients had one other visceral injury and 16 had at least two additional visceral injuries, including 26 hepatic, 24 splenic, and 16 renal injuries. Isolated adrenal hematoma in trauma patients is rare. Although nine patients had no other visceral injury aside from the adrenal hematoma, eight of these nine patients had substantial orthopedic, chest, or head injuries. We identified only one patient with adrenal hematoma as the only finding. In three cases, active contrast material extravasation indicative of active bleeding in the adrenal bed was detected. This is a very specific sign of injury, and like active extravasation from other organs, it is a sign of higher morbidity and mortality. One patient with active contrast material extravasation died within 24 hours of admission. The other two survived and were discharged after hospital stays of 30 and 40 days, which included, respectively, 24 and 33 days in the intensive care unit.
Although the exact mechanism of adrenal gland injury has not been proved, it is not surprising that these injuries occur in patients who have sustained more severe trauma, because the adrenal glands are small, relatively well-protected structures in the retroperitoneum. Sevitt (1) proposed two mechanisms of injury—severe hyperextension and lateral compressive force—either of which can result in direct injury that is probably due to compression against the spine. These direct injuries are often associated with other ipsilateral injuries: Left adrenal hematomas were more associated with left rib fracture and with splenic and left renal injuries in our study. Right adrenal hematomas were more common and were more highly associated with right rib fracture and with hepatic and right renal injuries. The greater propensity for right adrenal hematomas may be explained by the more confined space of the right adrenal gland between the liver and the spine and the greater mass of the liver.
Sevitt (1) postulated alternatively that the right adrenal gland may be particularly prone to injury owing to idiosyncrasies in the anatomy of this structure. The right adrenal gland, along with a very short direct adrenal vein, lies directly posterior to the inferior vena cava. With rapid deceleration injury and blunt trauma to the abdomen, very high abdominal pressure is probably transmitted through the vena cava, and thus through the adrenal vein, for a short duration. On the left, this pressure would dissipate through the left renal venous system, but on the right, it would be transmitted directly to the adrenal gland through the short right adrenal vein. In fact, pathologic examination of autopsy specimens has revealed disrupted venous walls with central (ie, medullary) hematomas (1). This mechanism could also explain the possibility of right adrenal hematoma in patients restrained by a combination of lap and shoulder belts and an airbag who experience a sudden deceleration with high abdominal pressure but avoid impact or hyperextension. The single case of an adrenal hematoma without any other substantial injury in our series was a right adrenal hematoma.
One potential difficulty in diagnosing trauma-induced adrenal hematoma is due to the common occurrence of incidental adrenal masses, particularly nonhyperfunctioning adenomas. The presence of a mass in the adrenal gland of a trauma patient should not be considered evidence of injury because it may be a preexisting incidental lesion. In this study, adrenal hematomas had the typical appearance: They had a mean maximum diameter of 2.8 cm, most were ovoid, and some were associated with periadrenal stranding. Most of the adrenal hematomas were slightly hyperattenuating, and they had a mean attenuation of 52 HU. Because trauma CT is usually performed after intravenous contrast material administration, there may be cases in which the distinction between adenoma and hematoma is not clear; in some of these cases, repeat nonenhanced scanning may be performed 24 hours later.
Determining the attenuation of lesions may be useful in distinguishing adrenal hematoma from incidental adenoma. Most acute adrenal hematomas are relatively hyperattenuating, while most adenomas are relatively hypoattenuating. In our series, the mean attenuation of the adrenal hematomas at initial CT was 52 HU. It was a technical limitation of our study that the initial CT images were obtained according to a trauma protocol; thus, imaging was performed only after intravenous administration of low-osmolar contrast material, with no precontrast images available to determine the nonenhanced attenuation. However, if one is uncertain about the type of lesion, images could be obtained after a 15-minute delay to assess for washout. Adenomas typically show substantial washout (12,13), whereas adrenal hematomas should show essentially no change in attenuation.
In most cases, delayed nonenhanced CT was not performed in patients; rather, follow-up CT, sometimes without contrast material enhancement, was performed if an adrenal mass was suspected. The adrenal hematomas seen at follow-up CT in this study usually had decreased in size and attenuation (28 of 34 lesions), showed resolution of periadrenal stranding, and, in some cases, had completely resolved. On the other hand, adenomas do not evolve. The distinction between hematoma (especially a unilateral hematoma) and adenoma is not clinically urgent and can be made at follow-up, particularly because many trauma patients undergo follow-up examinations.
This study had limitations that may have limited the conclusions that can be drawn: The study was retrospective in that it was based on a search of initial trauma CT reports. The CT images from all cases that were initially judged to be positive for adrenal hematoma were reviewed by two readers to confirm the diagnosis, and the CT images from cases that were initially judged to be positive for adrenal mass, which may or may not have been hematomas, were reviewed to make certain that all hematomas were included. The cases initially interpreted as negative, however, were not reviewed. Thus, theoretically, we could not accurately assess the sensitivity and specificity of CT for the detection of adrenal hematoma. It is theoretically possible that some adrenal hematomas were missed, and, thus, the corresponding images were incorrectly deemed to be negative, in this study. We believe that this is unlikely, however, because at our institution, all trauma CT images are initially read by a senior radiology resident and then overread; an experienced, board-certified body imaging specialist then signs the final report. In addition, all trauma CT images are also reviewed at a daily quality assurance conference that is attended by at least one and as many as six or more attending body imaging staff members. Because the normal appearance of the adrenal gland is easily recognized, images that are judged to be negative are very likely to be truly negative.
In conclusion, despite changes in the patterns of CT use, the prevalence of adrenal hemorrhage in trauma patients who undergo CT remains approximately 2%, but the frequency of detecting these lesions increases with increasing severity of trauma. Adrenal hematomas are associated with other injuries, especially hepatic, splenic, renal, chest, and orthopedic injuries. The subgroup of trauma patients who have adrenal hematomas also has a higher mortality rate. Although isolated adrenal hematomas may occur, the presence of an adrenal hematoma should trigger a careful search for other injuries. Conversely, if an adrenal mass is seen with no other injury in a patient who has sustained mild trauma, then an alternative diagnosis should be considered.
Adrenal hematomas have a typical appearance that can usually be recognized and distinguished from the appearance of incidental lesions. If necessary, CT images obtained 15 minutes after contrast material administration may demonstrate the washout that is typical of adenomas, and follow-up CT images may show the decreases in size and attenuation that are typical of hematomas.
|
APPENDIX: Injury Severity Score |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONAPPENDIX: Injury Severity ScoreREFERENCES |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, P.J.K.; study concepts, D.E.M.; study design, D.E.M., J.K.S.; literature research, P.J.K., M.E.L.; clinical studies, S.T.W., A.I.R.; data acquisition, A.I.R.; data analysis/interpretation, G.M., P.J.K.; statistical analysis, G.M.; manuscript preparation, P.J.K., A.I.R.; manuscript definition of intellectual content, P.J.K.; manuscript editing, M.E.L., J.K.S., G.M.; manuscript revision/review, P.J.K., M.E.L.; manuscript final version approval, P.J.K.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONAPPENDIX: Injury Severity ScoreREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. H. Song, F. S. Chaudhry, and W. W. Mayo-Smith The Incidental Adrenal Mass on CT: Prevalence of Adrenal Disease in 1,049 Consecutive Adrenal Masses in Patients with No Known Malignancy Am. J. Roentgenol., May 1, 2008; 190(5): 1163 - 1168. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
