We present the case of a patient who presented to our emergency department with isolated and severe head trauma by mechanism, but in whom abdominal imaging revealed an unexpected finding of an abnormal pancreas. The term pseudopancreatitis has been used in a trauma setting where there are CT findings of intra- and peripancreatic fluid (mimicking imaging findings of acute pancreatitis or low-grade pancreatic injury), but in the absence of direct abdominal trauma. To the best of our knowledge, there has only been one other case series, involving four patients, describing similar findings of isolated pancreatic abnormality in patients who have sustained trauma with no abdominal trauma [4]. Those patients, soldiers in the Israel-Lebanon conflict, sustained combat injuries to their extremities, but no abdominal injuries. They were treated with aggressive fluid resuscitation during their prolonged rescue time (mean time 4 hours 45 minutes to first CT). All these patients had decreased hemoglobin values and decreased hematocrit.
One theory postulated by the authors of that case series was overhydration secondary to massive or prolonged fluid resuscitation, where intra- and peripancreatic fluid likely represented extravasated saline and serum within the interstitium of the pancreas. However, by contrast, our patient was not resuscitated for a prolonged period of time before CT was performed, with the time from accident scene to CT scan totaling 1 hour and 15 minutes. Initial laboratory values in our emergency department (obtained just prior to the CT study) showed that our patient’s hemoglobin and hematocrit were within normal limits. The total fluid volume administered to our patient after initial blood work until the CT scan totaled no more than 3L, which included normal saline, packed red blood cells and mannitol. However, many trauma patients receive this volume of fluid during initial resuscitation, without pancreatic changes on CT. Also, whereas other signs of overhydration were noted on the CT studies of the patients in the case series - specifically, fluid tracking along the portal veins in the liver - this was not present in our patient.
In addition to overhydration, another possibility to account for the CT findings was pancreatic trauma. Traumatic pancreatic injury is uncommon, seen in 5% of (blunt) abdominal injury cases, and usually caused by direct trauma to the mid-to-upper abdomen [5]. An isolated pancreatic injury is even more unusual, as up to 70% of patients with pancreatic trauma have associated injuries, most commonly involving the duodenum, stomach, liver and spleen [3]. Depending on the mechanism of injury and the direction of force, the ascending colon, descending colon and kidneys may also be involved [6]. A diagnosis of pancreatic trauma is initially based on a combination of clinical findings, serum lipase or amylase levels, and radiological investigations. CT is currently the initial imaging modality of choice in stable patients, with reported sensitivities up to 85% in the first 24 hours of injury [6]. Endoscopic retrograde cholangiopancreatography (ERCP) and magnetic retrograde cholangiopancreatography (MRCP) can also play a role in diagnosing pancreatic duct injury. MRCP is useful in demonstrating peripancreatic fluid collections and pancreatic ductal anatomy, or disruption of the duct [7]. Secretin MRCP has been used as a problem-solving tool in pancreatic duct anatomy and integrity, and also to assess whether there is ongoing leakage from the pancreatic duct [7, 8]. ERCP can diagnose duct disruption and ongoing leaking, and can also be therapeutic in certain clinical scenarios, such as placement of a pancreatic duct stent [9]. However, ERCP is invasive and there is a risk of complications, including pancreatitis [9]. Despite the availability of these diagnostic tools, direct visualization via laparotomy may be required for a definitive diagnosis.
Although serum amylase and lipase are not always elevated in the setting of pancreatic trauma, up to 82% of patients with pancreatic injury will demonstrate an elevated amylase level [10]. In one study, where serial evaluation of serum amylase was performed, sensitivity increased to 90% [10]. Sensitivity of serum lipase in the adult population is not well known, but one study showed that 80% of patients with abdominal trauma demonstrated elevated lipase levels on initial investigation, with increasing levels seen on serial evaluation [10]. Our patient had an elevated serum amylase on blood work taken during her course in our intensive care unit, but her serum lipase was normal. However, serum amylase is not specific to pancreatitis, and can be seen in numerous other conditions, including cerebral trauma, which was present in this patient. As well, our patient’s serum lipase was normal, and given that the negative predictive value of serum lipase has been estimated to range from 94% to 100% [11], a diagnosis of traumatic pancreatic injury in our patient is unlikely. Thus, the pancreatic enlargement and peripancreatic fluid may have been due to another mechanism, such as systemic inflammatory response with cytokine production and release, perhaps related to the massive head trauma. These cytokines, in addition to complement system activation and other mediators, cause increased permeability of the vascular endothelium, which then results in tissue edema. These same mediators also alter cell membrane permeability, which can then cause cell swelling and rupture [12].
Pancreatic injuries are readily diagnosed at exploratory laparotomy [5]. Our patient, as an organ donor, underwent a laparotomy for organ harvesting. The transplant surgeon noted that her pancreas appeared edematous, but there was no evidence of traumatic injury to the organ, and no other signs indicative of trauma within the remainder of her abdominal cavity. The mechanism of injury and distribution of injuries sustained by our patient would make traumatic pancreatic injury highly unlikely. The history and physical examination suggested that a metallic object had struck our patient’s head directly, with no contact to any other part of her body, and nor did she experience blunt trauma from a steering wheel or front-end collision.
Another study documented peripancreatic fluid as a CT sign of shock in patients who sustained abdominal or pelvic trauma, in conjunction with clinical evidence of hypovolemic shock, but without evidence of pancreatic injury [13]. Shock pancreas is a well-documented entity, again seen mainly in the setting of hypovolemic shock, often due to traumatic injury [13–15]. It is thought that this is secondary to fluid shift across compartments, from the intravascular structures to the extravascular or extracellular spaces. This may be directly due to blood loss with replacement by hypo-osmolar fluids, or secondary to cytokine production and release, resulting in increased endothelial and cellular membrane permeability [13]. A similar entity, shock bowel, with marked mucosal enhancement and submucosal edema, is also usually seen in the setting of hypovolemic shock. However, it has now been described in patients who were not hypovolemic, or have not sustained significant blood loss. Shock bowel has been in documented in other clinical scenarios, including isolated head injury, sepsis and recent surgery [14]. It is possible that the pancreatic findings in our patient could represent shock pancreas secondary to neurogenic shock, although as far as we know, this particular entity has not been previously described. The significance of the elevated amylase but normal lipase level in our patient is uncertain. Some studies have documented normal pancreatic enzymes in patients with shock pancreas [13], while other patients demonstrate abnormally elevated amylase and lipase levels [16, 17].
Finally, other considerations for imaging findings would include acute pancreatitis secondary to other etiologies. Given that ultrasound is the modality of choice for gallstone imaging, no gallstones were seen on CT, nor was there any intra- or extrahepatic biliary duct dilation. As previously stated, given our patient’s serial lipase levels remained within normal limits, in conjunction with the high negative predictive value of serum lipase for acute pancreatitis, this is an unlikely consideration.