The spleen is after the liver as the second most frequently injured organ following abdominal trauma, occurring in 32% of abdominal injuries [1]. Splenic injuries most often are observed in blunt abdominal trauma, such as in motorcycle accidents, assaults, fall from height, and sports. During the past two decades, major changes in the management of splenic injuries have occurred. Traditionally, operative management (OM) was the standard for patients with splenic injury. Once the spleen has been mobilized, a decision must be made regarding splenectomy or splenic salvage procedures (mesh splenorrhaphy, partial resection, adhesive and/or coagulation techniques). Due to the increased risk of infections, in particular fatal overwhelming postsplenectomy sepsis, a trend from splenectomy toward splenic conservation has emerged [26]. Presently, nonoperative management (NOM) of splenic injury is the most common management strategy in hemodynamically stable patients [713]. NOM can be divided as observation (OBS) or angiography and embolization (AE). Observational management involves admission to a unit with monitoring of vital signs, strict bed rest, frequent monitoring of red blood cell count, and serial abdominal examinations [12, 14]. Improved imaging techniques and advances in interventional radiology have helped to differentiate patients who can be observed versus those who need AE. Nevertheless, the optimal patient selection is still a matter of debate and the role of CT and angio-embolization has not yet fully evolved.

The purpose of this article was to review the current literature pertaining to the diagnosis and transcatheter therapy of traumatic splenic injuries.



The primary goal in the initial management of abdominal trauma is to detect and treat life-threatening injuries, in the majority of cases bleeding-related, as quickly as possible. The method of choice for rapid evaluation of the abdomen for free fluid is the Focused Abdominal Sonography for Trauma (FAST). In European trauma centers, sonography has replaced diagnostic peritoneal lavage as the primary screening test for intra-abdominal hemorrhage. FAST can be performed simultaneously with resuscitation efforts during the initial trauma management and only takes 2 minutes to perform. For this reason, it also is useful in hemodynamically unstable patients [15]. FAST is especially useful for detecting the presence or absence of a hemoperitoneum, a herald of significant organ injury [1618] with a sensitivity of 90–93% [19]. However, FAST has a low sensitivity for detecting and grading splenic injury. Injuries to bowel and mesentery without hemoperitoneum and retroperitoneal hematomas also may be missed by FAST [20]. In addition to this, FAST is unable to detect the presence of active hemorrhage. For patients who are hemodynamically stable, a FAST detecting hemoperitoneum should lead to a CT scan for further evaluation of the nature and extent of injury.

When the FAST is negative for hemoperitoneum, debate exists regarding whether a CT scan is required. Estimates for the presence of intra-abdominal injuries in the absence of hemoperitoneum on FAST can be as high as 29% [16, 2123].

Most trauma centers that use FAST as a primary screening modality rely on the assumption that any missed injuries are low-grade lesions without serious clinical consequence [24, 25]. However, some studies demonstrated that the use of FAST examination as a screening tool for blunt abdominal injury in the hemodynamically stable trauma patient results in underdiagnoses of intra-abdominal injuries. Routine CT scan frequently reveals additional injuries, which results in a change of treatment in 6.4–16% of these patients. [20, 2628] This may have an impact on treatment and outcome in trauma patients. Therefore, hemodynamically stable patients with a negative FAST and a high clinical suspicion for splenic injury, for example, a seat belt sign or upper abdominal pain, should undergo routine CT scanning.

Computed tomography and grading systems

CT before and during the arterial and delayed phase of intravenous contrast enhancement is now considered the “gold standard” for diagnosing splenic injuries after trauma and is preferred in hemodynamically stable patients.

CT is the most accurate test to assess the grade of injury to the spleen, as well as other intraperitoneal and retroperitoneal organs, and it can give a relatively accurate estimation of the volume of hemoperitoneum. It also can detect the presence and location of active arterial hemorrhage as well as the presence of pseudoaneurysms or arteriovenous fistulas in the spleen [29, 30]. Active contrast extravasation is usually seen on CT scan as an irregular or linear area of contrast extravasation in the splenic parenchyma (patient A; Fig. 1), subcapsular space or in the peritoneum (patient B; Fig. 2) [28]. It may be difficult to accurately distinguish pseudoaneurysms from extravasation if monophasic scanning is performed during CT. If multiphasic CT is done, washout from the false aneurysm in the delayed phase enables reliable differentiation from extravasation [31].

Fig. 1
figure 1

Hemodynamically stable patient (patient A) with blunt abdominal trauma after fall from horse. CT with intravenous contrast shows small amount of hemoperitoneum around the spleen and a contrast “blush,” which is confined to the splenic parenchyma (AAST grade 3, Baltimore grade 4a)

Fig. 2
figure 2

Hemodynamically stable patient (patient B) with blunt abdominal trauma after motor vehicle accident. CT with intravenous contrast shows hemoperitoneum, fractured spleen with large hematoma and extravasation of contrast medium into the abdominal cavity (AAST grade 4, Baltimore grade 4b)

CT can play a role in hemodynamically unstable patients when the logistic situation is organized in such a way that CT scanning during initial trauma resuscitation and evaluation is directly available in or very close to the emergency room [32, 33]. Limitations of the early trauma CT include its modest sensitivity for assessing injury to pancreas, bowel and mesentery, as well as its relative inability to detect venous hemorrhage.

The most widely used grading system for blunt splenic injuries is the American Association for the Surgery of Trauma (AAST) injury scale [34, 35]. This grading system (Table 1) is based on the anatomic extension of disruption of the spleen, as shown on CT scans or during laparotomy. However, this grading system is not reliable for the prediction of the outcome of splenic injuries and not decisive about whether surgery of conservative treatment should be applied [2, 3639].

Table 1 Splenic trauma grading systems [34, 35, 45]

Previous studies have shown that the CT injury grade alone is a poor predictor for the success of NOM [36, 40, 41]. Recent literature has suggested that vascular injuries, including active splenic bleeding (contrast blush), pseudoaneurysms, and posttraumatic arteriovenous fistulas, are associated with an increased failure rate of nonoperative management [7, 10, 13, 28, 38, 4244]. However, these injuries are not included in the AAST grading system. Because of the risk of failure of nonoperative management of patients with these injuries on CT scan, it is important to identify these vascular injuries. Marmery and colleagues [45] have developed a new grading system (“Baltimore” grading system) (Table 1), which was based on experience from multiple trauma centers, indicating that CT evidence of a contrast blush or vascular injuries predicts the need for AE or surgical management [28, 38, 43, 46]. Patients with grade 4 injuries are candidates for splenic arteriography or splenic surgery. This “Baltimore” CT grading system seems to be better than the AAST system for predicting which patients are the most likely candidates for embolization or splenic surgery [45]. However, prospective, randomized studies are needed to validate these results.

Angiography and embolization


In 1995, Sclafani et al. described the first successful use of AE for hemodynamically stable patients with a splenic injury as adjunct to observational management [47]. Since the late 1990s, AE has been applied more frequently to achieve better splenic salvage rates for the treatment of patients with splenic injuries. Simple observation alone has been reported to have a failure rate as high as 34%; the rate is even higher among patients with high-grade splenic injuries (AAST grade III–V) [10, 37, 4851].

Many studies support the use of embolization as an adjunct to observation. AE has increased the success rate of nonoperative management by stopping ongoing bleeding as well as by preventing delayed rupture of the spleen [7, 47, 5257]. Success rates up to 97% are described in the literature. However, these results must be interpreted with some caution, because these results are based on cohort studies, which were compared with results of historical studies. Prospective, randomized, controlled trials have never been published.

Recent studies advocate the use of AE in the presence of the following CT findings: active contrast extravasation, pseudoaneurysm or arteriovenous fistula, large hemoperitoneum, and a higher grade of injury (grade III–V) [9, 10, 23, 28, 49, 55, 5862].

Several authors and the clinical practical guideline for the NOM of blunt splenic injury of the Eastern Association for the Surgery of Trauma suggest that the presence of a contrast blush or vascular injury on the CT scan may portend a higher failure rate after observational management [9, 10, 38, 43, 46].

Schurr et al. [46] described 89 patients who were initially managed nonoperatively. Twelve patients failed NOM. Upon review of the initial computed tomography scans, a contrast blush was noted in 8 of 12 (67%) patients who failed and in 5 of 77 (6%) of those who were successfully managed nonoperatively (p < 0.0001). Another study demonstrated that both a contrast blush and traumatic vascular injuries (pseudoaneurysms and arteriovenous fistula) were associated with a high failure rate (82%) of nonoperative management [43].

Omert et al. have questioned the significance of a contrast blush. They suggested that a contrast blush is not an absolute indication for an operative or angiographic intervention. Factors, such as patient age, grade of injury, and presence of hypotension, need to be considered in the clinical management of these patients [63].

Splenic embolization has been used successfully in patients with AAST grade 3–5 splenic injuries [9, 10, 55, 59, 64]. Haan et al. published results demonstrating success rates from 87–100% after embolization [9, 10]. Another study demonstrated increases in NOM success rate from 74–89% with embolization in patients with high-grade lesions [55].

Presence of a large hemoperitoneum predicted failure of NOM in the EAST study [49] and also in a study by Velmahos et al. [51]. However, in other studies a significant hemoperitoneum did not affect the success rate [9]. The combination of a contrast blush within the peritoneal cavity and a significant hemoperitoneum indicate active and massive bleeding and predict a high failure rate and thus warrant a low threshold for AE or a splenectomy [55, 61, 65].

The overall success rate (defined by preservation of the spleen) of splenic embolization ranges from 73–97% [7, 9, 10, 47, 54, 55, 57, 58, 61, 64, 66, 67], with most studies reporting success rates greater than 90%. Several authors also successfully re-embolized patients in whom initial embolization failed, which further increased the splenic salvage rate [9, 58, 67]. Some studies, in contrast, showed no benefit of splenic artery embolization. A study by Harbrecht et al. [68] concluded that patients who underwent splenic arteriography did not have improved nonoperative splenic salvage rates compared with a contemporaneous control group of similarly injured patients. In two other studies, the authors showed their concern that embolization may be overutilized for blunt splenic injury, resulting in still high failure rates (27%) after embolization [61, 69]. They recommend a low threshold to operate if bleeding persists in an embolized patient who had the combination of a high-grade injury and large hemoperitoneum or contrast blush on CT. Figure 3 depicts our strategy for the diagnostic and therapeutic management.

Fig. 3
figure 3

Flowchart for the diagnostic and therapeutic management of blunt splenic injury. *Abnormalities in physical examination of the abdomen, pelvis, or lumbar spine, base excess <−3, systolic blood pressure <90 mmHg, long bone fractures [27]


Splenic arterial catheterization is commonly performed using the common femoral artery access. Placement of a 5- to 6-French introducer sheath suffices in most cases. A flush aortogram may be performed to evaluate anatomy of the visceral vessels but is not mandatory, particularly when a contrast-enhanced CT scan is available. Selective angiography of the splenic artery should always be performed to evaluate arterial injury. Diagnostic series of the splenic artery should be obtained using a 4- or 5-French catheter, but for selective catheterization of splenic artery branches coaxial micro-catheters and micro-guidewires may be required. Techniques and materials used for embolization depend on anatomical considerations, the hemodynamic situation of the patient, and the type and distribution of vascular injuries. Coils are usually the embolic agents of choice, and gelfoam may in some situations be used. Other types of embolic agents are rarely needed. Occasionally, the use of an Amplatzer® vascular plug may be useful [70].

Occlusion after coil embolization usually occurs as a result of coil-induced thrombosis rather than mechanical occlusion of the lumen by the coil and therefore this technique works best when the coagulation profile of the patient is normal or only mildly abnormal. In case of serious clotting disturbances, addition of another embolic agent, such as gelfoam, is indicated. Gelfoam is a sterile gelatin sponge intended for use as a temporary intravascular embolic material. It can be used in the shape of a “torpedo” or as pledgets. It can be injected through both standard 4- to 5-French angiographic catheters and micro-catheter systems. The major advantage of coils compared with gelfoam is the ability to provide permanent embolization, which is most desirable in treating vascular injuries. Gelfoam is a biodegradable material, but splenic artery recanalization rates after embolization are not known. However, occlusive agents, such as gelfoam, seem to have a higher failure rate (50%) compared with coil embolization (23%) [61]. The routine use of antibiotic prophylaxis for splenic artery embolization is not indicated.

Proximal versus Selective Embolization

Both proximal splenic artery embolization (PSAE) and selective distal splenic artery embolization have been applied and are described in the literature [10, 47]. The surgical equivalent of PSAE for splenic injury—splenic artery ligation—was first described in 1979 [71]. PSAE (patient A; Figs. 4 and 5) is based on the theory that the intrasplenic blood flow and blood pressure decrease, as a result of which the bleeding stops [72]. Sufficient perfusion of the spleen for salvage of the organ is preserved through the collateral flow of the short gastric arteries. PSAE involves occluding the proximal splenic artery using coils or, in some specific situations, using the Amplatzer® vascular plug. It often can be performed using standard 4- to 5-French catheters and is usually less time-consuming than distal embolization [73, 74]. The technique of selective distal embolization (patient B; Figs. 6 and 7) involves embolizing only the injured blood vessels. This often involves smaller branches of the splenic artery, which are inside the spleen itself. Typically a micro-catheter system is required for this type of embolization, particularly when there is significant tortuosity of the splenic artery.

Fig. 4
figure 4

Same patient (patient A) as in Fig. 1. Selective digital subtraction angiogram of the celiac axis showing the intraparenchymal contrast “blush” in the spleen. Note tortuous anatomy of the splenic artery

Fig. 5
figure 5

Same patient (patient A) as Figs. 1 and 4. Selective embolization was not possible as a result of the tortuous anatomy and proximal coil embolization of the splenic artery was performed. Check angiogram after embolization shows occlusion of the proximal splenic artery. Perfusion of the spleen by collaterals is not shown in this image

Fig. 6
figure 6

Same patient (patient B) as in Fig. 2. Selective digital subtraction angiogram of the celiac axis showing multiple areas of contrast extravasation from peripheral branches of the splenic artery

Fig. 7
figure 7

Same patient (patient B) as in Figs. 2 and 6. Check angiogram after selective coil embolization of an interpolar branch of the splenic artery. Contrast extravasation is no longer seen, and there is good perfusion of the remainder of the spleen

Proximal embolization is mostly used when there is diffuse bleeding of the spleen, when there are multiple focal bleeding vessels in the spleen, when there is time-pressure as a result of the hemodynamic situation of the patient, or when tortuosity of the splenic artery prevents selective distal embolization. PSAE also may used in some situations where the site of hemorrhage is not identified on the angiogram and the clinical situation of the patient suggests ongoing bleeding form the spleen.

Selective distal embolization is usually reserved for patients who have one or only a few focal bleeding vessels in the spleen and in whom the anatomy and hemodynamic situation allow employment of this.

The current literature provides little evidence whether PSAE or selective embolization is a better treatment. Although no prospective studies comparing the results of proximal or selective embolization are described, the use of PSAE seems to have some advantages. PSAE is faster, is associated with a lower failure rate of NOM in some studies, and a lower incidence of splenic abscess or infarction compared with selective distal embolization [64, 75]. However, it has to be noted that splenic infarcts after selective distal embolization rarely lead to clinical sequelae and splenic abscess can usually be treated percutaneously. Several studies have shown that PSAE and selective embolization have no major long-term impact on the splenic anatomy and immune function [42, 76, 77].

A potential disadvantage of PSAE could be the fact that in case of rebleeding, repeat embolization is difficult due to the inaccessibility of the splenic artery. Nevertheless, both techniques have been used successfully. The choice for either of the two techniques depends on the considerations mentioned above as well as operator preference.


The lifetime risk of overwhelming postsplenectomy sepsis is 1–2%, with a mortality rate of 33%. For that reason, the current recommendation is to treat patients after splenectomy with H. influenzae (Hib), pneumococcal, and meningococcal vaccination [78]. Several studies have shown that proximal ligation of the splenic artery and proximal and selective splenic embolization has no major long-term impact on the splenic anatomy and immune function [42, 76, 77]. Only one recent study suggests that the immunologic profile of embolized patients is reduced to controls. These studies, however, state that large controlled studies will be needed to make definitive vaccination recommendations [79]. We applied vaccination in patients after extensive distal embolization or embolization (proximal or distal) for a shattered spleen.


For effective application of AE as a treatment modality, CT scanning should be available 24 h per day to triage patients between observation, angio-embolization, and surgery. Early assessment using CT scanning in the emergency room may further improve these logistics [80, 81]. Also, the hospital interventional radiology suite and personnel should be set up for rapid response at any time.

The success rate of NOM depends on the time between the initial intake at the emergency room and the AE, as a result of the decreasing clinical condition and coagulation state of the patient [82]. Because embolization can be time-consuming and there is a risk for hemodynamic deterioration, the patient should be monitored carefully. Therefore, AE requires good teamwork among the trauma surgeon, the anesthesiologist, and the interventional radiologist.


Major complications after splenic artery embolization have been reported in 6–20% [9, 10, 83]. In a small study by Ekeh and colleagues [69], major complications, such as bleeding, abscess, and contrast nephropathy, occurred in 27%. Minor complications, fever, pleural effusions, and coil migration occurred in 53%. Other studies have demonstrated lower complication rates [9, 10, 83]. In the Western Trauma Association multi-institutional trial [9], 140 patients underwent proximal and distal embolization and combination of both techniques in 83, 48, and 9 patients, respectively. Mean AAST splenic injury grade was 3.5. The following vascular injuries were seen on the CT scan: hemoperitoneum (59%), contrast extravasation (44%), pseudoaneurysm (33%), and AV fistula (4%). This trial reported a major complication rate of 20%. These complications included persistent bleeding or rebleeding (11%), missed injury (3%), and splenic abscess (4%). Two percent of the patients had coil migration at the time of angiography, without sequelae in any of these patients. Infarctions after embolization occurred in 21% of patients in this study. Similar infarction rates have been described in other series [9, 61, 75, 83]. In one study, the infarction rate was 100% after distal embolization and 63% after proximal embolization [66]. Most patients with splenic infarcts are asymptomatic and could be managed conservatively [61, 66]. Abscess usually presents at contrast-enhanced CT as a hypodens fluid collection with or without air or an air-fluid level. However, the presence of gas in the spleen after embolization is not a specific sign for infection. In general, postembolization splenic abscesses can be treated with percutaneous drainage. Puncture site-related complications (hematoma, dissection, thrombosis) after angiography are relatively rare [47, 66].


AE is widely accepted for the management of hemodynamically stable patients with splenic injury, due to the improved CT techniques, their direct availability in early trauma management, and the advances in interventional radiology. CT scanning with IV contrast aids in selecting hemodynamically stable patients for AE, but the optimal patient selection is still controversial. The single CT finding that warrants immediate AE is a contrast blush within the peritoneal cavity, particularly for patients who are not hemodynamically stable or who have other clinical signs of ongoing bleeding. Other CT features, such as false aneurysm, AV fistula, contrast extravasation contained within the spleen, and the presence of a hemoperitoneum, are associated with an increased failure rate of NOM, but whether angio-embolization for such findings increase the success rate of NOM remains a matter of debate.

Angio-embolization has shown to be a valuable adjunct to observational management and has increased the success rate of NOM in many series. PSAE and selective embolization can both be performed and there is currently no evidence favoring either technique. However, these results must be interpreted with caution, because they are based on cohort studies, which were compared with results of historical studies. Prospective, randomized, controlled trials comparing observational management versus embolization have never been published. Although difficult to conduct due to the nature of the trauma population, such trials are needed to further determine the optimal patient selection for AE for traumatic splenic injury.