Analytical review of 664 cases of penetrating buttock trauma

A comprehensive review of data has not yet been provided as penetrating injury to the buttock is not a common condition accounting for 2-3% of all penetrating injuries. The aim of the study is to provide the as yet lacking analytical review of the literature on penetrating trauma to the buttock, with appraisal of characteristics, features, outcomes, and patterns of major injuries. Based on these results we will provide an algorithm. Using a set of terms we searched the databases Pub Med, EMBASE, Cochran, and CINAHL for articles published in English between 1970 and 2010. We analysed cumulative data from prospective and retrospective studies, and case reports. The literature search revealed 36 relevant articles containing data on 664 patients. There was no grade A evidence found. The injury population mostly consists of young males (95.4%) with a high proportion missile injury (75.9%). Bleeding was found to be the key problem which mostly occurs from internal injury and results in shock in 10%. Overall mortality is 2.9% with significant adverse impact of visceral or vascular injury and shock (P < 0.001). The major injury pattern significantly varies between shot and stab injury with small bowel, colon, or rectum injuries leading in shot wounds, whilst vascular injury leads in stab wounds (P < 0.01). Laparotomy was required in 26.9% of patients. Wound infection, sepsis or multiorgan failure, small bowel fistula, ileus, rebleeding, focal neurologic deficit, and urinary tract infection were the most common complications. Sharp differences in injury pattern endorse an algorithm for differential therapy of penetrating buttock trauma. In conclusion, penetrating buttock trauma should be regarded as a life-threatening injury with impact beyond the pelvis until proven otherwise.


Background
The buttock comprises the lateral half of the lower most sagittal zone of the torso [1] where there is a particularly high density of vital structures above and below the peritoneum in the pelvis [2,3]. Sparse evidence points to the frequency of life-threatening visceral and vascular injuries in patients with penetrating trauma to the buttock [2,4,5]. Pelvic anatomy results in the possibility of major complications or death following penetrating buttock injury in any path of trajectory and in absence of hard vascular, abdominal, or pelvic signs [4].
A comprehensive review of data has not yet been provided as penetrating injury to the buttock is not a common condition accounting for 2-3% of all penetrating injuries [3,[6][7][8][9][10]. Four previous reviews of the literature do however require additional research in terms of consistent patterns, peculiarities, and management [6][7][8][9].
The purpose of this study is to provide an analytical review of the literature on penetrating trauma to the buttock and to appraise the characteristics, features, outcomes, and patterns of major injuries. Recognition of specific patterns should enhance management of this trauma.

Methods
The Entrez PubMed interface of MEDLINE database, EMBASE, Cochran, and CINAHL databases were searched using the following Medical Subject Heading (MeSH) keywords: "Injuries", "Wounds and Injuries", "Wound Penetrating"; each of these keywords was combined with the keyword "Buttocks". The term 'Penetrating Gluteal Injuries' was also used. This resulted in 1021 titles and abstracts of studies related to these terms which were then read on the basis of English language and relevance.
Commentaries and literature reviews were also taken into account. We excluded articles relating to blunt injury, acupuncture injury, intragluteal injection injury, needle stick accidents, iatrogenic injury of the gluteal arteries, wound closure, reconstructive surgery of gluteal defects, wound botulism, bone fracture complications, injury from ultraviolet light, burn injury, true aneurisms, malignancies, and animal studies.
Relevant studies on penetrating buttock injury in acute trauma setting were grouped and categorised chronologically. Clustered and individual data regarding the demographic characteristics, mechanism of injury, clinical mode of presentation, imaging, buttock zone wounded, injuries, management strategy, complications, and final outcome were accumulated from all the studies, either prospective or retrospective, and case reports. When calculations in main series were impossible due to the lack of particular data, they were performed through the use of informative subset with indication of the exact number of entered cases.
In order to assess outcomes of visceral, vascular, skeletal, nerve injuries as well as outcomes of major surgery after stabbing or shootings, the 95% confidence intervals of odds ratios were calculated. In order to detect differences in injury related with stabbing or shooting patterns and outcomes between two independent proportions a Z-test was chosen and employed as both sample sizes were greater than 30. The two-tailed test was used to assess the null hypothesis. Chi-square test with Yates' correction was employed to compare categorical "alive -dead" outcome. Two-tailed p values were calculated where by P < 0.05 was considered to indicate statistical significance. Microsoft Office XP Excel 2007 Worksheets were used for accumulation and analysis of data.
Clinical presentation on admission was known in 654 patients. 74 patients (11.3%) were regarded haemodynamically unstable and 56 (8.6%) were diagnosed to be in haemorrhagic shock. Peritoneal irritation was present in 48 (7.3%), gross rectal blood in 41 (6.3%), and gross haematuria in 27 (4.1%) patients. Massive external bleeding was documented in 15 patients, false aneurysm formation in 12, absence of distal pulse or cold painful leg in two, groin hematoma in two, and severe bone pain in three patients. Initial diagnostic procedures were described by the authors as follows: diagnostic proctosigmoidoscopy in 295 (45.1%), angiography in 47 (7.2%), urology imaging (cystography, intravenous pyelography, urethrography) in 27 (4.1%) patients, and CT-scan for 10 (1.5%) patients. Retrograde irigoscopy and diagnostic peritoneal lavage were mentioned in a few reports.

Treatment modalities
The treatment approaches were described in 654 patients. 176 (26.9%) patients underwent emergency laparotomy. 40 (6.1%) patients required extended gluteal surgery. The interventional radiology procedures were used as sole modality to control bleeding or target bullets in 12 patients (1.8%). 356 (54.4%) patients were observed without major procedure. Other surgical procedures such as debridement under general anaesthesia were performed in 16.5% (n = 108) of patients.

Outcomes Mortality
Overall mortality rate was 2.9% (19/664). In terms of stabbing injury the mortality rate was 3.8% (6/158) and Table 1 Major endpoints of two prospective [11,12] and twelve retrospective reviews on penetrating buttock injury in acute trauma setting Clinical examination is a reliable predictor Maull et al. [13] ( 2.6% (13/504) following missile injuries. Mortality rate due to gunshot injuries was 2.2% (10/457). 6.4% (3/47) of patients admitted for blast injuries had died. Both patients treated for impalement survived. Details related to each fatality due to penetrating injuries to the buttock are demonstrated in Table 2. Hypovolaemic shock, major surgical intervention, and visceral and/or vascular injury are all factors which have a significant impact on a lethal outcome (Table 3).
The details of major injuries due to penetrating trauma to the buttock is shown in Figure 1. 30 anatomical terms were used to describe a particular injury type. The small bowel (8.3%), colon (6.3%), superior gluteal artery (5.4%), rectum (4.9%), bony pelvis (4.4%), bladder (3.7%), and iliac artery (2.0%) were on the top of the drawing scale of damaged anatomical structures. Summing up data on large bowel and major junctional vessel injury demonstrated that prevalence of injury to large bowel was 11.2% (n = 69); it was 2.9% for iliac artery or vein injury (n = 18), and 1.3% (n = 8) for femoral artery or vein injury. 10 major vessels injured due to penetrating buttock trauma were not named. Gluteal arteries were damaged in 37 patients (6.0%).
Pattern of major injuries related to shot wounds 225 major injuries were identified in the subset of 457 patients with gunshot injury (Figure 3). There were 166 visceral injuries (36.3%), 27 injuries to the bony pelvis  The pattern of major injury relating to injury mechanism Table 4 demonstrates a higher frequency for all visceral and skeletal pelvic injuries in the patients with shot wounds. Injuries to the organs located more distally from the wound site (colon, small bowel, and bladder) were far more frequently damaged in patients with shot wounds to the buttock. Rectum and major vessels of the region (iliac vessels, femoral vessels, and gluteal arteries) were damaged more frequently in patients with stab wounds to the buttock.

Penetrating injuries to the upper vs lower zone of the buttock
A subset including 97 cases from two retrospective studies [3,17] and six case reports [21,22,25,27,29] provided data to assigns the main wound site to the upper or lower buttock region. Statistical results regarding penetrating injuries above and below the intertrochanteric line are shown in Table 5

Discussion
It may be helpful to remind ourselves of the former surgical perspective on buttock trauma. Feigenberg (1992) reviewed four papers on stab wounds to the buttock and concluded that any stab wound to this body region should be regarded as potentially dangerous and every effort should be made to locate possible injuries [6]. Salim and Velmahos' review (2002) on abdominal gunshot wounds contains only one chapter regarding injury to the buttocks [7] and refers to one reference [11] pointing out that haemodynamically stable patients should be triaged (operation vs adjunct investigations) according to findings of physical examination. Aydin (2007) highlighted the importance of placing an acute  false aneurysm in the differential diagnosis of an indurate, fluctuant, warm, erythematous posttraumatic gluteal mass [8]. The key statements of the review provided by Butt (2009) [9] are based on the summary of three papers [11,12,37] on gunshot wounds to the buttocks, back, and pelvis: firstly, the management of gunshot wounds of the buttocks should follow the same principles with anterior abdomen gunshot wounds; secondly, clinical examination is a reliable predictor for the need of an operation; thirdly, a rigid sigmoidoscopy is introduced per routine for all patients. Case reports on penetrating buttock injury [6,8,[19][20][21][22][23][24][25][26][27][28][29][30][31][32][33] highlight the importance of a thorough and aggressive evaluation of the patient [6], observation [23,27], prompt differential diagnosis [8,21,30,31], immediate assessment of the lower urinary tract [21,22], and lately the value of dynamic 2D and 3D CT-scanning and angiography [28]. They also highlight rare complications following highvelocity or low-velocity gunshot injury to the buttock where the bullet or pellet migrates to major veins such as inferior cava vein and hepatic veins [29] or if it reaches the right ventricle of the heart [23], needing a broad range of approaches ranging from open surgery to angioembolization [6,21,22], transjugular extraction of bullet from middle hepatic vein [29], image navigation surgery [33], gluteal surgery [28,32], laparoscopy [24], and laparotomy [6,20,21,25].
Our analytical review demonstrates that penetrating trauma to the buttock is a serious diagnostic and clinical concern with a mortality rate of 2.9%. Mortality of penetrating stab injuries to the buttock is comparable to that of extra-buttock regions of the body, such as penetrating injury to the posterior abdomen is 0-2% [37][38][39], the anterior abdomen 0-4.4% [40][41][42][43], the thoracoabdominal area 2.1% [44], and the chest 2.5-5.6% [44][45][46]. Mortality may be less in cohorts with isolated stab injury to the chest (1.46%) [45], or after exclusion of cardiac injuries (0.8%) [44]. Regarding pelvic or transpelvic gunshot trauma, mortality rates vary from 0-12.2% [11,47,48]. Cohorts with gunshot wounds to the limbs may show no mortality [49,50]. We conclude that penetrating injuries to the buttock poses a similar threat to the patient as penetrating trauma of any other body region.
Despite the fact that stab wound primarily cause locoregional damage, whilst gunshot trauma is associated with frequent extraterritorial injury, stab wounds (3.8% mortality rate) are even more dangerous than missile wounds per se or gunshot wounds specifically (2.6% and 2.2% mortality rate, respectively). Injury of buttock due to impalement remains uncommon [26,51]. It is therefore recommended to classify impalement related injuries as a separate category of penetrating injuries [52].
Analysis of the associated major injuries due to penetrating trauma to the buttock reveals several unexpected particularities. The most commonly damaged particular organs and vessels were, in descending order, small bowel, colon, superior gluteal artery, and rectum. Injury of iliac artery and/or vein was a rare, but relevant finding with 2.9%. This counterintuitive finding is better understood on analysis of subgroups created according to injury mechanism.
As expected, stabbings were most frequently associated with injuries to gluteal arteries (20.3%), rectum (19.0%), and iliac vessels (8.2%). The prevalence of injuries to femoral artery or vein was 3.8%. Gunshot injuries frequently result in wider organ damage involving small  Table 4 provides ample evidence that gunshot and stab trauma of the buttock are actually two separate clinical entities. They require different diagnostic and surgical approaches which are summarised in Figure 4. In our view, such an approach based on empiric evidence might usefully supersede former algorithms by trying to address particular aspects of buttock trauma [2,5,14,17]. This review confirms the conclusion of two other authors [3,17] suggesting that injuries of upper zone of the buttock are associated with higher probability of viscus or major vessel injury comparing with injuries to the lower zone of the buttock. Table 5 reveals significant differentiation of injury patterns according to zone of primary injury site. However, the low positive predictive value does not recommend to rely on this criterion, for management strategies based on division of the buttock. On any account, the frequency of extraregional injury should prompt an aggressive and speedy computed tomography imaging approach to the entire abdomen and pelvis, complemented by a chest x-ray in all gunshot wounds to the buttock.
The current review contains a significant amount of historical data, bringing the use of endovascular approaches to only 1.8% in the current cohort. The advent of interventional radiological techniques should enable embolisation of pelvic vessels beside the level of the common or external iliac vessels [36,53].
Selective non-operative management of penetrating trauma to the buttock in stable patients without evidence of major organ injury is a successful approach [11]. Serial clinical examination should include per rectal examination, rigid sigmoidoscopy, and urinanalysis because of quite high probability of colorectal (11.2%) as well as bladder, urethra, and ureter injury (5.4%).
A classification of CT findings into three main groups of subset in relation to stable patients (abdominal/pelvis injury, gluteal vessel injury, and femoral vessel injury) is another feature of the algorithm (Figure 4). The rationale of this is the following: the buttocks should be regarded as a distinct anatomical/junctional zone in trauma surgery because patterns of penetrating injury and clinical characteristics as well as implications of buttock trauma disclosed in this paper correspond with general hallmarks of junctional trauma [54].  In terms of injury severity score, only Ferraro [16] and Lesperance [10] used the ISS scale. It is important to emphasise coding technique for penetrating buttock injury according to newest AIS 2005 © Update 2008 [55]. It indicates that superficial (minor) penetrating injury to the buttock should be regarded as grade 1 (code 816011.1). When there is tissue loss >25 cm 2 , it should be regarded as grade 2 injury (code 816012.2), and when it is associated with blood loss >20% by volume, it has to be regarded as grade 3 injury (816013.3). Such injuries should be assigned to the external body region when calculating the ISS. However, if underlying anatomical structures are involved, documented diagnoses should be coded only, and they should be assigned to either the lower extremity body region or abdomen. Penetrating injuries involving a bone is coded as open fracture to the specific bone.
There are several limitations of this review. Publication bias, retrospective approach, clustered data, complexity of some injuries, and constrained nature of this study are the factors which undoubtedly cause our bias views. Prospective networked studies would be a better approach to the problem. The current review may help to design such studies.
In conclusion, penetrating buttock trauma should be regarded as a life-threatening injury with impact beyond the pelvis until proven otherwise.