Laparoscopic splenectomy as a definitive management option for high-grade traumatic splenic injury when non operative management is not feasible or failed: a 5-year experience from a level one trauma center with minimally invasive surgery expertise

The spleen is the most commonly injured solid organ in abdominal trauma, and the most commonly injured structure in the abdomen following blunt trauma [1]. Despite the extensive use of non-operative management (NOM) and newer adjuncts such as angioembolization (AE) for hemodynamically non-compromized patients with splenic injuries even in high-grade injuries [2], splenectomy in hemodynamically non-compromized or ‘quasi-stable’ patients continues to play an important role in trauma surgery [3].

Although there are clear benefits to NOM (with or without AE) in terms of avoidance of splenectomy and complications associated with a laparotomy, there are many potential negative factors associated with NOM. These include its costs and morbidities, it is not always technically feasible or successful, it requires a strict patient conduct and close expert monitoring, it does not completely prevent delayed splenic rupture or hemorrhage, and it requires the immediate availability of an operating room and operative team at all times [4].

Indications for splenectomy in hemodynamically non-compromised patients include blunt or penetrating splenic injury requiring surgical exploration for diaphragmatic or hollow-viscus injuries, high-grade blunt splenic injury with unavailable, contraindicated, unfeasible or unsuccessful NOM and AE and all complications following AE such as pseudoaneurysms, splenic infarction or abscess, and delayed rupture. NOM is contraindicated in patients unable or unwilling to comply with the strict NOM conduct and activity restrictions (e.g., mentally impaired, homeless, self-employed, professional athletes), as well as those with an unreliable examination typically due to associated injuries and intubation [5‐9]. There are also contraindications to AE as an adjunct, including patients severely allergic to intravenous iodine contrast or with late stage chronic kidney disease. AE may also be unfeasible because of a non-cooperative patient or due to technical reasons, such as celiac trunk stenosis, tortuous and kinking splenic artery, or failure in releasing the coils. In some cases, even if AE is technically performed, it can fail to achieve hemostasis or cause splenic infarction and abscess formation that requires percutaneous interventional or operative interventional. Furthermore, due to the high risk of NOM failure (80%), some authors recommend splenectomy in grade IV splenic injuries with sub capsular hematoma or vascular abnormalities (pseudoaneurysms) [10]. Splenectomy may be performed in patients with blunt splenic injury and multiple severe skeletal injuries requiring prolonged and invasive orthopaedic procedures in a prone position to avoid the risk of simultaneous bleeding from multiple sites. According to the surgeon’s evaluation, splenectomy may also be indicated in patients affected by splenomegaly (e.g., lymphoma, tropical, auto-immunity, portal hypertension [6, 11]) or in any case the trauma surgeon deems NOM and AE not safe or indicated. Table 1 summarizes the possible indications for splenectomy in hemodynamically non-compromised trauma patients.

As a large proportion of the morbidity associated with standard open splenectomy is attributed to the incision (pain, respiratory distress, wound infection, incisional hernia) and the development of adhesions from open bowel manipulation, the use of minimally invasive techniques represents an intriguing potential option to treat the splenic injury while avoiding many of the risks associated with open surgery. In addition, minimally invasive surgery allows for a faster post-operative recovery and shorter length of stay (LOS) compared to open surgery [12, 13]. Therefore, laparoscopic splenectomy (LS) may be an advantageous alternative to open splenectomy (OS), either as an early or delayed procedure for splenic trauma, failure of nonoperative management, or treatment of AE associated complications [14, 15]. Advantages and disadvantages of NOM and LS are listed in Table 2. Nevertheless, LS in hemodynamically non-compromised patients with splenic injuries is not widely accepted yet and only a few case reports and small case-series have been published in the literature [16‐25].

In this article we sought to analyze our institutional experience with LS for splenic trauma in hemodynamically non-compromised patients. Our primary objectives were to investigate the indications, safety, feasibility, and outcomes of LS in this group of patients. In addition, we sought to outline key tips and tricks for performing LS for trauma, as well as describing the step-by-step surgical technique and an associated procedural video.

All patients underwent contrast-enhanced computed tomography (CE CT-scan) on admission as part of their radiologic trauma evaluation. Decision to consider NOM not feasible or failed and indications for a definitive splenectomy was made by experienced trauma surgeons in a multidisciplinary discussion with Trauma ICU and Interventional Radiology attendings (Consultants).

All patients were hemodynamically stable and fully resuscitated, and all had a High-Grade (≥ 3) splenic injury. Selection between the open or laparoscopic approach was made randomly and exclusively based upon the calendar of the on call roster and the availability on duty of an Attending Trauma Surgeon with advanced minimally invasive surgery (MIS) expertise. Whenever the on call Attending Trauma Surgeon was a surgeon without advanced MIS expertise, the splenectomy was made via a traditional open midline laparotomy. Whenever was on duty the on call attending experienced trauma surgeon with advanced laparoscopic skills and a completed formal training (fellowship or above) in MIS (SDS fulfilled these criteria in our institution), the splenectomy was approached/performed laparoscopically.

In either case, the criteria for the Decision to consider NOM not feasible or failed and indications for a definitive splenectomy did not differ and was made by experienced trauma surgeons in a multidisciplinary discussion with Trauma ICU attending and Interventional Radiology Attendings. Selection of the open or laparoscopic approach between the 2 group was made exclusively based upon the calendar of the on call roster and the availability on duty of an Attending Trauma Surgeon with advanced MIS expertise.

Exclusion criteria for enrollment in the study were:

Early procedure was defined as a surgical operation performed within 24 h from admission, while this was defined as delayed if performed after 24 h from admission.

Failed splenic NOM (with or without AE) was defined as:

Between January 2013 and December 2017, 48 hemodynamically stable patients underwent splenectomy for splenic trauma at our institution: 32 procedures (67%) were OS, while 16 of them (33%) were approached in a minimally invasive manner (LS), with a conversion rate of 19% (3/16) (Table 3). All but 1 patient were blunt trauma.

Both the intention-to-treat and per-protocol analyses (Tables 3, 4) of the overall population did not show significant differences between the groups baseline characteristics in terms of age, vital signs, laboratory values, ASA score, ISS, AAST injury grade, and associated injuries. NOM with AE had an overall failure rate of 50% among the 16 patients in the LS group. This included one early AE failure (unfeasible due to celiac trunk stenosis) and 7 cases with initial success of AE at haemorrhage control but then later failure due the development of pseudoaneurisms in 4 patients, oozing in 2 patients and splenic abscess in 1 patient. NOM with AE had been successfully performed but later failed in 10 out of 32 cases in the OS group (31%) (oozing in 5 patients, delayed rupture in three patients, splenic infarction in 2 patients). Indications for splenectomy were decided upon the multidisciplinary evaluation but also by the on-call trauma surgeon given his/her evaluation, experience and discretion upon clinical evaluation, imaging characteristics, and laboratory values at presentation and over time (Table 5). Early procedures were more common in the OS group (71%, 23/32), while delayed procedures were more common in the LS group (56%, 9/16). The most delayed procedure was 15 days from trauma. The average delay was 6 days in the open group and 2 days in the laparoscopic group. In examining the operative and post-operative outcomes, there were several statistically significant differences in favour of LS. These included an increased incidence of associated surgical procedures performed (p 0.016) and a significantly shorter time to bowel function recovery compared to OS (p < 0.0001). The operating time (p < 0.0001) and ICU length of stay (p 0.042) were significantly higher in the LS group, but these differences appeared to be mainly attributed to the increased incidence of associated procedures in the LS group and not to the splenectomy itself. The 2 techniques were comparable in terms of morbidity, mortality, overall LOS and long-term complications. Since a significantly higher rate of associated surgical procedures were found in the LS group, further intention-to-treat and per-protocol analyses for the subgroup of isolated splenic injuries were performed (Tables 6, 7). This sub-analysis demonstrated no significant difference in ICU length of stay between the 2 groups and demonstrated similar findings as described for the entire cohort. Of note, although operative time with LS was significantly longer vs OS, the laparoscopic approach was associated with a significantly faster recovery in terms of time to NGT removal, time to initiate diet, and time to first bowel movement.

The cornerstone of debate about splenectomy in stable high-grade splenic injury (AAST grade III–V) is the role of NOM and AE. Recent meta-analysis by Crichton et al. including 23 studies demonstrated how AE allows for decrease in rate of NOM failure from 45–75% to 9–12% in grade IV–V splenic injuries [26]. Similar results were reported by Requarth et al. in 2011 [27]. However, according to the study by Dolejs et al. on the national trauma database between 2008 and 2014 [28], splenectomy rate seems to be stable despite the increased use of AE regardless of the grade of splenic injury, questioning if this should be the first choice treatment in high-grade stable splenic injuries [3].

Hypotension in the field and on initial presentation to the ED, large volume blood transfusion, altered mental status, age above 55 years, hemoperitoneum volume > 250 ml, higher ISS/AAST-OIS grade, associated abdominal injuries, and splenic vascular abnormalities (pseudoaneurysms, arteriovenous fistulae or contrast blushes) have all been identified as risk factors for NOM failure [28‐31]. Similarly, factors including subcapsular hematoma and intra-peritoneal contrast blush on CT scan have been suggested as additional independent risk factors for requiring splenectomy [6, 32]. NOM (± AE) failures after penetrating splenic injuries in stable patients can be particularly common, due to the frequent high grade of splenic trauma and the high rate of diaphragmatic lacerations (up to 60%) and other visceral injuries that require immediate laparotomy [33].

Another critical but often underappreciated concern with AE is the rate of major early and delayed complications, including re-bleeding, splenic infarction, abscess, acute renal insufficiency, ARDS, femoral pseudoaneurysms and other access-site complications [28, 33‐41]. Although AE is often touted as a “spleen preserving” adjunct to NOM that helps preserve the innate immune function of the spleen, recent literature is calling this assumption into question. In an analysis of 37,986 patients from the Nationwide Readmissions Database, AE was associated with a significantly increased risk of both early and late infectious complications vs NOM without AE, and even had a higher rate of organ space infection at 1 year compared to patients who underwent splenectomy [42]. Despite the current discussions on the usefulness of AE, the risks and complications related to the procedure, and the risk factors for failure, all hemodynamically non-compromised patients with splenic injury, where NOM with or without AE fails or develops complications can be efficiently managed by LS.

NOM (± AE) for splenic injuries has some clinical, social and economical costs as well that must be taken into consideration. In fact, it requires close monitoring, reliable patient adherence to instructions and strict conduct for several weeks, long-term follow-up to reduce the risk of delayed rupture, as well as repeated evaluations or imaging to rule-out possible complications (e.g., pseudoaneurysms) [10, 43]. Therefore, not all patients may be suitable for this kind of management strategy. For instance, fit adult patients or professional athletes may prefer to have the spleen removed instead of facing a long hospital stay and many days of home recovery and strict conduct limitations in order not to lose business commitments. Other examples of patients, where NOM (± AE) may not be indicated are mentally impaired or socially disadvantaged and vulnerable individuals (e.g., homeless).

In addition, general surgeons outside major trauma centers may not feel comfortable in pursuing NOM in higher splenic injury grades and with the spreading of minimally invasive skills some may feel more confident in performing a laparoscopic splenectomy than a high-risk NOM.

Furthermore, to safely undertake NOM (± AE) for high-grade splenic injuries, an operating room and operative team immediately available at all times is required, but this is not the case in many hospitals.

Moreover, there are conditions, where AE fails because of technical difficulties (such as celiac trunk stenosis or tortuous kinking of the splenic artery) which precludes adequate or safe embolization and haemorrhage control. Another major complication of AE is splenic infarction and subsequent conversion to a splenic abscess that requires percutaneous drainage or operative intervention. AE also does not exclude the risk of delayed splenic rupture, and when this occurs there is often no choice but to proceed with emergent open splenectomy.

Some polytrauma patients may require urgent prone spine surgery or prolonged orthopaedic surgery with significant risk of bleeding [6, 43]. In other cases, diaphragmatic injury is associated with splenic trauma and needs immediate repair. In addition, NOM in penetrating splenic injury is rarely indicated. Finally, in patients with splenic injury in pathologic splenomegaly (lymphoma, tropical, autoimmunity, etc.) splenectomy might be considered the treatment of choice. All these examples are good candidates for LS. Other cases are deemed candidate to splenectomy solely based on the trauma surgeon’s evaluation and experience.

Literature on LS for trauma is overall limited, but the published case reports and case-series all suggest that the laparoscopic technique is safe and feasible [14, 17‐24]. In 2015 Ermolov et al. published a case series of 23 LS, showing how these patients had better recovering conditions when compared to laparotomy without increased complications or mortality [19]. It was acknowledged that the minimally invasive procedures were more time-consuming. More recently, Shamin et al. reported on the national trauma database (113 laparoscopic splenectomies) demonstrating the safety and feasibility of the procedure in selected cases and experienced hands [15] in terms of comparable mortality, length of stay and major complications, but did not investigate on post-operative bowel recovery. Both these published experiences found similar results to the present study.

Unique from other case series, our current study includes 1 case of penetrating trauma. The results of this study demonstrate how LS in hemodynamically stable trauma patients can be equally indicated in both young and elderly patients, low and high AAST grade splenic injury, low and high ISS score. Compared to OS, LS significantly improves the post-operative recovery with no increased morbidity or mortality. According to these results and in line with the previous published data [14‐24], LS seems non-inferior to OS in selected hemodynamically stable trauma patients, where NOM has failed or is deemed not-indicated. However, the presence of both adequate laparoscopic skills and experience in trauma surgery are fundamental to undertake this kind of procedures and achieve good results. The longer operative time of LS has been reported in the other similar publications and is expected to decrease along with improving surgeon’s experience and confidence.

The main strength of the study is the single operator laparoscopic case series that allowed for uniform operative data. Secondly, the high specialization of the institution (level I trauma center), with a dedicated multidisciplinary trauma team (radiologist, intensivist, surgeon) allowed a high standard of trauma care and expertise in dealing with trauma patients. On the other hand, a crucial limitation of the study is the small sample size, which can be explained with patient selection, the novelty of the technique and the low overall volume of abdominal trauma in European countries. The small sample size of our study probably affects the generalization applicability of the results but is proportionate to the national trauma database report by Shamin et al., where 113 LS were performed across the USA in 9 years (12 per year in the whole USA) [15] which is, to the best of our knowledge, the largest single operator case series of LS for trauma. Another limitation is the retrospective analysis that prevented complete reporting on factors such as postoperative pain. Finally, there was inter-operator variability with our laparoscopic experience being solely performed by a single surgeon and our open splenectomy cases having multiple trauma surgeons contribute to our experience.

A potential criticism to a more liberal policy for enlarging the indications to a definitive laparoscopic splenectomy to the patients who are stable but have high-grade splenic injuries (grade III and above ± contrast blushes) is the loss of the spleen and its immunologic and hematologic function, including the theoretical increased risk of overwhelming post-splenectomy infection (OPSI). Nevertheless the immunologic function of the spleen is clinically significant mainly in paediatric patients and young ages. OPSI incidence may be over-estimated and may be very low, even in paediatric populations. In a recent study including 116 children with BSI, 27 underwent splenectomy and 66 patients were treated by a spleen preserving therapy (including embolization) [44]. Only 2 out of 27 splenectomized patients were adequately vaccinated, 5 patients without a spleen used prophylactic antibiotics, and about half of the asplenic patients had adequate knowledge of the risk that asplenia entails. A total of 22/27 splenectomized patients were neither adequately vaccinated nor received prophylactic antibiotics. There was no OPSI seen in this study population during the 1116 follow-up years [10, 44]. Incidence of OPSI in adults is reported in isolated cases than true epidemiological data. The precise incidence of OPSI remains controversial. Overall, the most reliable data related to incidence estimate approximately 1 case occurring per 500 person-years of observation. Asplenic children younger than 5 years, especially infants splenectomized for trauma, may have an infection rate of greater than 10% [43, 45]. Physicians should also be careful not to interpret as an OPSI, the occurrence of severe chest or abdominal infections or other infections or seasonal flu that are NOT or unlikely to be related to splenectomy and would have occurred regardless the patient had their spleen removed.

Finally, laparoscopic splenectomy can easily be combined with autotransplantation of a fragment or single segment of the native spleen, even if this was shattered, and may, therefore, represent a minimally invasive possibility for preserving splenic function after traumatic splenectomy [46, 47]. Laparoscopic Autotransplantation of a fragment of the native spleen after morcellation and extraction of the traumatized spleen, is theoretically worth practicing and likely the subject of a future prospective study.

Laparoscopic splenectomy for trauma patients is gradually gaining acceptance among the scientific community. Due to the variable nature of trauma, only small case series with low scientific evidence exists [48]. Nevertheless, all published data and the present study demonstrates that laparoscopic splenectomy in hemodynamically non-compromized patients with splenic injuries not amenable to NOM (or failed NOM) and adjuncts such as angioembolization is safe and feasible regardless of the age of patients, severity of trauma, or presence of associated injuries. LS is associated with non-inferior morbidity and mortality and significantly improved post-operative recovery compared to standard open splenectomy; however, the potential advantages and safety of minimally invasive surgery must be considered in relation to the level of expertise of the institution, the availability of adequate laparoscopic equipment, and most importantly the presence of an experienced and skilled laparoscopic surgeon. Prospective or randomized controlled trials in patients with hemodynamically non-compromized or ‘quasi-stable’ splenic injuries are needed to better investigate this cutting-edge topic and better characterize the short and long term benefits of laparoscopic splenectomy for trauma.