Pancreatic and duodenal injury is rare, occurring in less than 1% of all trauma admissions [1]. It has however been reported to have a high morbidity (36-60%) and mortality (18-23%) [6]. The low incidence of the pancreaticoduodenal injuries is thought to be due to its position in the retroperitoneum, shielded somewhat from the anterior abdominal wall and more anterior structures [4]. This may have implications of delayed diagnosis and is usually complicated by other intra-abdominal injuries [6-8]. Grading of pancreatic trauma is determined by the location of the injury and the presence of ductal damage [6]. Duodenal trauma is graded by injury thickness, extent of circumference of the lacerated lumen, and involvement of the common bile duct or ampulla [6].
Management of these injuries depend on the degree of injury. Clinical assessment and prompt diagnosis is important to minimise morbidity. A normal serum amylase has previously been reported to occur in up to 40% of patients with pancreatic trauma [1]. In our cohort, it was normal in 35% of patients, however this was limited by six patients that did not have a serum amylase or lipase measurement on admission. Having a baseline level for later comparison may still be of some value [1].
Cross-sectional imaging is necessary and surgical intervention may be required [1]. Contrast enhanced CT has a high specificity (90-95%) but low sensitivity for pancreatic ductal involvement (52-54%) [1]. Magnetic resonance imaging pancreatography (MRCP) can help with the diagnosis of ductal injuries [9]. Endoscopic retrograde pancreatography (ERP) can also identify pancreatic duct injuries, however, is more invasive and has an associated risk of morbidity.
MRCP should be considered for detection of pancreatic duct injuries when a pancreatic injury is suspected on cross sectional CT imaging (sensitivity of 90-100%) [10]. ERCP has the additional benefit of allowing endoscopic treatment of a ductal injury, but this must be balanced by the associated risk of ERCP related morbidity [11]. Moreover, distal pancreatectomy is still considered the gold-standard treatment for pancreatic transection with ductal injury. ERCP and pancreatic stent placement can also be of benefit for management of complications of pancreatic injuries such as pseudocysts [11-13]. Five patients in our cohort had an MRCP to exclude ductal injuries, of which two patients were managed non-operatively after.
It should be noted that even in patients with Grade I pancreatic injuries, other organs injuries were common in our cohort. As the pancreas sits in a retroperitoneal position, other intra-abdominal injuries complicated management and recovery [6-8]. For our cohort, 12/45 were managed non-operatively. All but one of those patients (11/45) had grade I or II injuries according to AAST grade. Several studies have described the role of endoscopic management for pancreatic duct injury, negating the need for operative management. However, it is worth noting that most of these studies were performed in the paediatric population. One patient from our cohort with a ductal injury was managed with an ERCP and pancreatic duct stent. Two patients had ERCP to manage complications from their pancreatic injuries (duct stricture + fistula and pseudocyst), one of which, had a late presentation after being readmitted with fevers and abdominal pain. This pancreatic duct fistula (distal tail) and duct stenosis (main body) was successfully managed with an endoscopic pancreatic stent placement. Grade IV and V injuries generally require operative management, with pancreaticoduodenectomy (Whipple’s procedure) being indicated if there is a massive disruption of pancreatic head [12,13].
Treatment for grade I and II duodenal haematoma can be managed non-operatively, but duodenal lacerations will require surgical repair [9,10] . Duodenal obstruction may occur with large mural haematoma formation [4]. The majority of grade II lacerations can be repaired by primary repair [4] [Fig. 8], which in our cohort, all were. One patient was managed with primary repair and malecot drain insertion for decompression of D3. If the duodenum is unable to be repaired, then a pancreaticoduodenectomy (Whipples’ procedure) may be indicated [4]. Weale et al noted in their retrospective study that 91 patients out of 94 had a primary repair [3]. Only three patients in their cohort had pyloric exclusion. However, the majority of patients in their series were injured from a penetrating mechanism and nearly all had AAST grade II injuries [3]. In comparison, in our series of predominantly blunt trauma, ten patients had primary closure and four of which had a pyloric exclusion, all with grade III injuries. In addition, one patient from our cohort had a side-to-side duodenojejunostomy.
Complication rates for pancreatic trauma are variable and are reported to range from 26-86%, depending on severity [12,13]. The most common complication cited is a pancreatic fistula (10-35%) [1,10] which can be managed with drains [11-13], although persistent fistulas may benefit from endoscopic pancreatic duct stenting [12]. Other complications include post-traumatic pancreatitis, intra-abdominal abscess, and pseudocyst formation [1,11-13]. Three patients (10%) in our cohort with pancreatic injuries developed pancreatic collections which were drained. Only two patients in our cohort developed pancreatic fistula (7%).
Søreide et al [10] noted the risk of mortality with grade I pancreatic injuries with no other injuries was < 5% and if in shock with associated other injuries is still < 10% [10]. Grade IV and V injuries with associated shock and other injuries have an associated morbidity risk of > 50% and mortality risk 20-50% [10]. Our overall mortality rate was 3 patients out of 45 (7%), and only two of these patients died as a result of complications attributable to pancreaticoduodenal injury (4%). Krig et al [2] have noted in that in their retrospective study of 473 patients – mortality rate was 15% and that deaths, whilst uncommon, occur late and due to multiorgan failure and sepsis which is in keeping with our data [2].
This study is limited by its retrospective data and that pancreatic and duodenal injuries are rare, even in trauma centres, and thus the actual number of cases were low (45 cases in 13 years). Three patients were transferred from regions outside of our ability to follow them up, thus it is possible that late complications in these three patients were missed.