The Flat Cava Sign Revisited: What is its Significance in Patients Without Trauma?

Transcription

1 Recha S. Eisenstat 1 Allen C. Whitford 2 Michael J. Lane 2 Douglas S. Katz 1 Received June 14, 2001; accepted after revision August 2, The opinions and assertions contained herein are the private views of the authors and are not to be construed as official or as representing the views of the Department of the Army. Presented at the annual meeting of the American Roentgen Ray Society, Seattle, April May Department of Radiology, Winthrop University Hospital, 259 First St., Mineola, NY Address correspondence to D. S. Katz. 2 Brooke Army Medical Center, 3851 Roger Brooke Dr., Fort Sam Houston, San Antonio, TX AJR 2002;178: X/02/ American Roentgen Ray Society The Flat Cava Sign Revisited: What is its Significance in Patients Without Trauma? OBJECTIVE. The purpose of this study was to evaluate the clinical significance of the flat cava sign on abdominal CT scans in hospitalized patients without trauma. MATERIALS AND METHODS. CT scans of the abdomen of 500 inpatients imaged for a wide variety of nontraumatic indications were retrospectively reviewed for a flat cava sign. Two radiologists measured the maximal anteroposterior and transverse diameters of the inferior vena cava at four predetermined levels. The medical records of the subset of patients with a flat cava sign defined as a maximal transverse-to-anteroposterior ratio of 3:1 or greater at one or more of the four levels were reviewed for evidence of hypovolemia or hypotension. RESULTS. Seventy patients (14%; 48 women, 22 men) had a flat inferior vena cava present on at least one of the four levels. Of these 70 patients, 21 had definite and three had possible clinical evidence of hypotension or hypovolemia. A flat cava sign isolated to only one level was seen in 22 of the 70 patients, most commonly at the level just below the renal veins, and only four of these 22 patients had evidence of hypotension or hypovolemia. CONCLUSION. Of the 500 inpatients, 14% had a flat cava sign on at least one of the four levels examined on abdominal CT scans. The majority of these patients with a flat cava sign did not have hypotension or evidence of hypovolemia, but a minority (30%) did. P revious studies have suggested that the presence of a flattened inferior vena cava is an indicator of hypovolemia or hypotension in patients who undergo CT of the abdomen and pelvis after blunt abdominal trauma and that the identification of such a flat cava sign may signify impending cardiovascular collapse [1 4]. In 1987, Jeffrey and Federle [1] retrospectively reviewed the abdominal CT scans of 100 patients with blunt abdominal trauma. There was a flat cava sign, which was defined as flattening of the inferior vena cava on at least three contiguous 1-cm images, in seven patients. All seven patients were believed to be stable enough to undergo CT; the severity of these patients hemodynamic compromise was not realized immediately before CT, although after CT, six of the seven had major hemorrhage that required surgery and three patients subsequently died of their injuries. Of the control group of 100 patients in this study who underwent abdominal CT for nonacute conditions, 98 had a round or oval inferior vena cava, two had flattening on less than three consecutive images, and none had a flattened cava on more than three images. Jeffrey and Federle concluded that the flat cava sign at multiple levels in trauma patients is a sign of hypovolemia due to major hemorrhage. However, in a study by Rak et al. [5], the caliber of the infrahepatic inferior vena cava varied at sonography in normal volunteers, depending on the phase of respiration and on differences in intraabdominal pressure. Similar findings were present in a sonographic study by Grant et al. [6]. We anecdotally noted the presence of the flat cava sign in occasional patients without trauma who underwent CT of the abdomen, especially elderly women, and wondered if the sign had any relationship to low blood pressure or hypovolemia or if it was purely a variation of normal. To our knowledge, the flat cava sign has not been systematically studied on abdominal CT scans in nontrauma patients. Therefore, we conducted a retrospective study to evaluate the significance of the flat cava sign on abdominal CT scans in patients without trauma. Materials and Methods A retrospective review was made of 500 nearly consecutive abdominal CT scans obtained on a single hospital-based CT scanner (HiSpeed Advantage; AJR:178, January

2 Eisenstat et al. General Electric Medical Systems, Milwaukee, WI). A few outpatients were not included in the series so that the 500 patients were all inpatients who were imaged for a wide variety of indications. CT scans obtained for abdominal or pelvic trauma were also excluded, as were repeated abdominal CT scans so that the 500 studies were acquired in 500 different patients. In general, most of the examinations were performed with oral and IV contrast, and the slice thickness was 7 mm. One of two radiologists, both of whom were unaware of the clinical information, reloaded the CT scans onto a workstation. The maximal transverse (relative to the long axis of the cava on axial images) and anteroposterior dimensions of the inferior vena cava were measured using electronic calipers at four predetermined levels: just below the intrahepatic cava, just below the renal veins, midway between the renal veins and the caval bifurcation, and just above the caval bifurcation. To establish a reproducible cutoff, we recorded an inferior vena cava as flattened if the ratio of the maximal transverse-to-maximal anteroposterior dimension was 3:1 or greater at one or more of the four levels and as not flattened if the ratio was less than 3:1. The total craniocaudal extent of the flattened portion of the inferior vena cava (whether contiguous or not) was also estimated in centimeters on the basis of review of the individual axial images. If the inferior vena cava was flattened at any of the four levels, the exact date and time that CT had begun was recorded. After review of the 500 CT cases, the medical records of those patients with a flattened inferior vena cava on at least one of the four levels were reviewed in detail. The medical records were evaluated for evidence of hypotension or hypovolemia within a 12-hr time window spanning 6 hr before and 6 hr after CT, but particular attention was paid to the medical record relating to the hour around the exact time of the CT. Specific criteria used for determining that hypotension was present included a systolic blood pressure less than 100 mm Hg, especially if associated with a heart rate of greater than 140 beats per minute. Data on vital signs during this 12- hr window were available for all patients. Specific criteria used for determining that hypovolemia was present included evidence of fluid depletion at physical examination, such as dry mucous membranes and poor skin tone, and an analysis of fluid inputs and outputs including requirements for fluids and blood. The medical records were also carefully searched for improvement in blood pressure and volume status after treatment. Results A flattened inferior vena cava, defined by a maximal transverse-to-maximal anteroposterior diameter of 3:1 or more, was found on 70 (14%) of the 500 CT scans, on at least one of the four predetermined levels. Of these 70 patients, 21 (30%) had definite evidence of hypovolemia or hypotension (Fig. 1), three (4%) had possible evidence of hypovolemia or hypotension, and 46 (66%) had no evidence of hypovolemia or hypotension (Fig. 2). Of the 21 patients with definite evidence of hypovolemia or hypotension, 17 had a systolic blood pressure of less than 100 mm Hg; 11 had clinical signs of hypovolemia (i.e., dehydration), including seven of the 17 patients who were also hypotensive. Two of the 21 patients were being administered dopamine to maintain blood pressure and systemic perfusion, and three required multiple blood transfusions to correct ongoing blood loss. The primary clinical diagnoses in these patients at the time of or immediately after the abdominal CT included retroperitoneal hemorrhage, sepsis, dehydration, bowel ischemia, cholangitis, colitis, anemia, abdominal abscess, acute myocardial infarction, and pericardial effusion and tamponade. The clinical and CT findings are summarized in Figure 3. The three patients with equivocal findings of hypotension or hypovolemia on chart review had histories as follows: the first, a 78- year-old woman with sepsis and a normal blood pressure had tachycardia, which might have been purely a result of sepsis, although hypovolemia could not be excluded. The second, a 77-year-old woman, had diarrhea and anemia and was normotensive (blood pressure 133/70 mm Hg at the time of CT). However, she was receiving IV fluids, and her inputs were significantly ahead of her outputs for a 48-hr period. The third patient, a 90-year-old woman with anasarca, anemia, and equivocal findings for urosepsis, initially had a blood pressure of 148/50 mm Hg at the time of CT, but this decreased to 108/40 mm Hg within 6 hr after CT. A The total number of males in this study was 235, and the total number of females was 265. The age range of the 500 patients was 2 98 years. Of the 21 children 18 years or younger in the study (4% of the total), none had a flattened inferior vena cava at any level. Of the 70 patients with a flat cava sign on CT scans, there were 48 women and 22 men; this difference was found to be statistically significant (p < 0.01, Fischer s exact test, two-tailed). There were seven men and 14 women composing the group of 21 patients with definite hypotension or hypovolemia around the time of CT, and 15 men and 31 women composing the group of 46 patients who did not have hypotension or hypovolemia. The mean age of the 21 patients with hypotension or hypovolemia was 65 years, and the mean age of the 46 patients without hypotension or hypovolemia was 69 years. The mean age of all 70 patients with a flattened inferior vena cava on at least one level at CT was 68 years, and the mean age of the 430 patients without a flattened inferior vena cava at any level was 57 years; this age difference of 11 years was found to be statistically significant (p < , Wilcoxon s rank sum test). Of the 21 patients with hypotension or hypovolemia, a flattened inferior vena cava was present at two or more of the predetermined levels in 17. There was flattening at two levels in 10 patients, at three levels in five patients, and at four levels in two patients. Of the 46 patients who were not hypotensive or hypovolemic, a flattened inferior vena cava was present at two or more levels in 28. There was flattening Fig year-old man with hypotension at time of CT scan and flat cava sign at first three of four levels. Patient had severe colitis, which led to sepsis, hypovolemia, and hypotension (blood pressure, 100/50 mm Hg) and was being administered dopamine to maintain blood pressure. Total craniocaudal extent of flattened inferior vena cava was approximately 10 cm. A, CT scan with oral and IV contrast shows flattening of inferior vena cava (arrows) just below intrahepatic cava. Ratio of maximal transverse (2.8-cm) to anteroposterior (0.6-cm) dimension of cava was 4.7:1. B, CT scan at level just below renal veins shows flattened inferior vena cava (arrow). Ratio of maximal transverse (2.7-cm) to anteroposterior (0.6-cm) dimension of cava is 4.5:1. B 22 AJR:178, January 2002

3 Significance of the Flat Cava Sign at CT at two levels in 25 patients (including three patients with discontinuous flattening at the first and third levels), at three levels in two patients, and at four levels in one patient. Of the three patients with equivocal findings of hypovolemia or hypotension, there was flattening of the inferior vena cava at two levels in two patients and at three levels in the other patient. A flattened inferior vena cava was seen at only one of the four levels in 22 of the 70 patients. In nine of these 22 patients, the flattened inferior vena cava was isolated to the level just below the renal veins; eight of these nine patients were women. None of these nine individuals were hypotensive or hypovolemic. Of the other 13 patients, there were seven with isolated flattening of the inferior vena cava at the level just below the liver, five with flattening only at the level halfway between the renal veins and the caval bifurcation, and one with flattening only just above the caval bifurcation. Of these 13 patients, only four had evidence of hypotension or hypovolemia. Mean measurement of the total craniocaudal extent of the flattened cava in the 21 patients with definite hypotension or hypovolemia, estimated to the nearest centimeter, was 7 cm, whereas this mean distance in the 46 patients without hypotension or hypovolemia was 5 cm. Discussion The association of hypovolemia and hypotension in patients with a flattened inferior vena cava on abdominal CT scans was initially noted in the late 1980s in blunt trauma patients [1, 2]. Jeffrey and Federle [1] reported a strong correlation of hypovolemia with flattening of A Fig year-old woman with flattened inferior vena cava at first three of four levels and normal blood pressure. Patient was imaged for epigastric pain and was found to have sigmoid diverticulitis. Craniocaudal extent of flattened inferior vena cava was approximately 10 cm. A, CT scan with oral contrast shows flattened inferior vena cava (arrows) at level just below intrahepatic cava. Ratio of maximal transverse (3.0-cm)-to-anteroposterior (0.5-cm) dimension was 6:1. B, CT scan at level just below renal veins shows flattened inferior vena cava (arrows), with ratio of maximal transverse (3.6-cm)-to-anteroposterior (0.6-cm) dimension of 6:1. C, CT scan at level halfway between renal veins and caval bifurcation shows flattened inferior vena cava (arrows), with ratio of maximal transverse (2.7-cm)-to-anteroposterior (0.7-cm) dimension of 3.9:1. the inferior vena cava at multiple levels on CT scans in patients who had suffered from substantial blunt abdominal trauma. Retrospective review of the abdominal CT scans of 100 trauma patients in the study by Jeffrey and Federle revealed flattening on at least three consecutive 1-cm-thick images in seven patients. Six of the seven patients had experienced major hemorrhage. The severity of the hemodynamic insult was not initially realized in five of these patients before CT, and three patients subsequently died. A control group of 100 abdominal CT scans obtained in patients B Fig. 3. Summary of clinical status and inferior vena cava flattening on CT scans in 500 inpatients without trauma. without acute conditions showed a round or oval inferior vena cava in 98 patients, flattening of the cava on fewer than three consecutive images in two patients, and flattening of the inferior vena cava on more than three consecutive images in no patients. Taylor et al. [2] concluded that a flat cava sign may be a sign of impending cardiovascular collapse. More recently, Mirvis et al. [4] noted the presence of a flattened inferior vena cava on abdominal CT in 10 of 13 patients who were clinically in shock after blunt abdominal trauma. They defined a flattened cava as less than 9 mm in the C AJR:178, January

4 Eisenstat et al. maximal anteroposterior dimension at the level of the renal veins. The flattened inferior vena cava has also been identified on abdominal CT scans in pediatric trauma patients and has been closely associated with hypotension and hypovolemia. Taylor et al. [2] reported this CT finding in three patients less than 2 years old, after severe abdominal trauma. They described a hypoperfusion complex in these patients that, in addition to a flattened inferior vena cava, included decreased caliber of the aorta, marked diffuse bowel distention with fluid, moderate to large peritoneal fluid collections, and abnormally intense enhancement of the bowel wall, kidneys, and pancreas. The three children in this report required initial aggressive fluid resuscitation but appeared to be stable enough at the time that the decision was made for an abdominal CT scan to be obtained. However, all three children died of their injuries [2]. Similarly, Sivit et al. [3] reported the hypoperfusion complex in 27 of 1,018 children who underwent contrast-enhanced CT of the abdomen after blunt trauma. Flattening of the inferior vena cava, seen over at least 3 cm in the craniocaudad dimension, was present in all of these 27 children [3]. Most recently, O Hara and Donnelly [7] described the CT findings of the hypoperfusion complex in six pediatric trauma victims and speculated that inferior vena caval flattening is due to a combination of hypovolemia and vasospasm from increased sympathetic activity related to partially compensated shock. To our knowledge, the flat cava sign has not been previously studied on abdominal CT scans in patients without blunt abdominal trauma to determine if it has the same significance as it does in trauma patients. On the basis of anecdotal cases observed before this study, we initially believed that a flat cava sign did not correlate with either hypotension or hypovolemia in any patient. Similarly, Hopper [8] anecdotally noted occasional multilevel collapse of the inferior vena cava on abdominal CT scans in nontrauma patients and speculated that this finding was most commonly a normal variant that depended on the degree of respiration, intraabdominal pressure, and hydration status. In our retrospective study of 500 inpatients who did not have blunt abdominal trauma but who did have a wide variety of other emergent and semiurgent indications for abdominal CT, the majority who had a flat cava sign on at least one of the four levels that we examined (46 patients) in fact did not have evidence of hypovolemia or hypotension around the time of CT. However, a minority of patients (21/70 with a flat cava sign) did have a definite correlation with hypotension or hypovolemia, and three had a possible correlation. Therefore, in 24 patients (34%) with a flat cava sign, this finding was significant; it, therefore, cannot be dismissed solely as a normal variant of older nontrauma patients. Although the mean craniocaudal extent of the flattened cava was slightly longer (7 cm) in the patients with hypotension or hypovolemia compared with the patients without either of these findings (5 cm), the presence of this finding at only one of the four levels that were specifically examined in this study was much more likely to be identified in normotensive and normovolemic patients; only four of the 22 patients with a flat cava sign isolated to one level in this study had hypotension or hypovolemia. There is a variety of potential explanations for the flattened inferior vena cava that we identified in our normovolemic and normotensive adult patients. These include a normal variation, especially in elderly women; a change in the vessel tone or connective tissues within the caval wall that occurs with aging; a redistribution of blood volume that does not manifest as clinical hypotension or hypovolemia; and variations in caval shape and volume with ventilation, intraabdominal pressure, and position of the patient. To our knowledge, none of these potential explanations have been investigated with CT, but there have been two sonographic studies that addressed some of the potential explanations. Rak et al. [5] prospectively evaluated the infrahepatic inferior vena cava of 26 normal volunteers on sonography and showed significant variation in caval caliber related to ventilation and intraabdominal pressure. Similarly, Grant et al. [6] studied the inferior vena cava in 25 normal volunteers at sonography. During inspiration and Valsalva s maneuvers, the inferior vena cava significantly decreased, whereas it significantly increased during expiration and breath-holding [6]. There are some potential limitations of our study. This study was retrospective, and although we were initially unaware of the clinical history when we performed the image review, we were aware that all patients whose medical records we examined did have a flattened inferior vena cava. We did not examine the medical records of the 430 patients without a flattened inferior vena cava to determine the incidence of hypotension or hypovolemia, but investigating this group was not the purpose of our study. Also, we chose a maximal transverse-to-maximal anteroposterior caval ratio of three or greater as the criteria for a flattened cava; although this may have been somewhat arbitrary and without specific precedent in the relatively limited radiology literature on this topic, we believed that an objective cutoff was necessary, rather than merely declaring that a cava was flat or not flat. There was little guidance in the previous literature, other than a single paper which defined flattening of the inferior vena cava as 9 mm or less (in the anteroposterior diameter), measured at the level of the renal veins [4]; all other studies simply stated that the inferior vena cava was considered diminished if it appeared flattened, typically on multiple or at least three contiguous images [1 3, 7]. A ratio may be more reproducible than a single dimensional measurement and may account for differences in body habitus. It is unclear, however, how many more patients with hypotension or hypovolemia would have been identified if a higher or lower ratio had been used as a threshold; presumably, the specificity of a flattened cava sign for hypotension or hypovolemia would have increased or decreased, respectively, while its sensitivity would have decreased, or increased, respectively. Additionally, a minority of the patients in this study did not received IV contrast for CT, but the relatively small volume of fluid that was administered during most of the CT examinations probably would not have made a significant difference in caval status. More vigorous fluid resuscitation as opposed to a small volume of fluid would likely be needed to have a significant effect on central venous pressure and volume [9]. Furthermore, some elderly patients may have a relatively low blood pressure as a baseline, so the clinical significance of hypotension is not always clear. Finally, although we attempted to determine the volume status of patients with a flat cava sign through a detailed chart review, a prospective study would have allowed more systematic evaluation of each patient s volume status; ideally such an evaluation would also include central venous pressure monitoring, although this obviously is not possible in most patients. In summary, of the 500 nontrauma inpatients in our study who underwent abdominal CT, 14% had a flat cava sign. The flattened inferior vena cava was more commonly seen in women and in older patients. When isolated to one of the four levels examined in this study, particularly at the level just below the renal veins, the flat cava sign most often occurred in elderly women who did not have evidence of hypotension or hypovolemia. In contrast to patients with blunt trauma, the presence of a flattened inferior vena cava was more common in patients without hypotension or hypovolemia, but 30% of patients 24 AJR:178, January 2002

5 Significance of the Flat Cava Sign at CT did have definite evidence of hypovolemia or hypotension. The findings of a flat cava sign should, therefore, not be ignored but closely correlated with the clinical findings. Further prospective studies in both trauma and nontrauma patients would help clarify the exact significance of the flat cava sign on abdominal CT scans. Acknowledgments We thank the biostatistical department and CT technologists at our institution for their assistance with this project. References 1. Jeffrey RB, Federle MP. The collapsed inferior vena cava: CT evidence of hypovolemia. AJR 1988;150: Taylor GA, Fallat ME, Eichelberger MR. Hypovolemic shock in children: abdominal CT manifestations. Radiology 1987;164: Sivit CJ, Taylor GA, Bulas DI, Kushner DC, Potter BM, Eichelberger MR. Posttraumatic shock in children: CT findings associated with hemodynamic instability. Radiology 1992;182: Mirvis SE, Shanmuganathan K, Erb R. Diffuse small-bowel ischemia in hypotensive adults after blunt trauma (shock bowel): CT findings and clinical significance. AJR 1994;163: Rak KM, Hopper KD, Tyler HN. The slit infrahepatic IVC: pathologic entity or normal variant? J Clin Ultrasound 1991;19: Grant E, Rendano F, Sevinc E, Gammelgarrd J, Holm HH, Grφnvall S. Normal inferior vena cava: caliber changes observed by dynamic ultrasound. AJR 1980;135: O Hara SM, Donnelly LF. Intense contrast enhancement of the adrenal glands: another abdominal CT finding associated with the hypoperfusion complex in children. AJR 1999;173: Hopper KD. The slit inferior vena cava. (letter) AJR 1988;151: Shanmuganathan K, Mirvis SE, Amoroso M. Periportal low density on CT in patients with blunt trauma: association with elevated venous pressure. AJR 1993;160: AJR:178, January