- Email: [email protected]
The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery
G Model JINJ 7527 No. of Pages 7
Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery Beatrice Sannaa,b , Brandon Michael Henrya,c,* , Jens Viksea,d , Bendik Skinningsruda,c , Jakub R. Pe˛kalaa,c, Jerzy A. Walochaa,c, Roberto Cirocchie , Krzysztof A. Tomaszewskia,c a
International Evidence-Based Anatomy Working Group, Krakow, Poland Faculty of Medicine & Surgery, University of Cagliari, Sardinia, Italy Department of Anatomy, Jagiellonian University Medical College, Krakow, Poland d Division of Medicine, Stavanger University Hospital, Stavanger, Norway e Department of Surgical Sciences, University of Perugia, Italy b c
A R T I C L E I N F O
Keywords: Corona mortis Obturator vessels Hernia repair Pelvic lymphadenectomy Retropubic region Pelvic fracture Vascular anatomy
A B S T R A C T
Purpose: Corona mortis is a highly variable vascular connection between the obturator and external iliac or inferior epigastric arteries or veins located behind the superior pubic ramus in the retropubic space (space of Retzius). Due to the significant variation in this collateral circulation, detailed anatomical knowledge of the corona mortis is vital to enhance the prevention of possible iatrogenic errors in hernia repair and other pubic surgical procedures. The aim of our meta-analysis was to provide comprehensive data on the prevalence, anatomical characteristics, and ethnic variations of the corona mortis vessel. Methods: An extensive search was conducted through the major electronic databases to identify eligible articles. Data extracted included investigative method, prevalence of the corona mortis vessels among hemi-pelvises (overall, arterial only, venous only, and combined), distance from the corona mortis to pubic symphysis, and assessment of gender, side, laterality, and ethnicity subgroups. Results: A total of 21 studies (n = 2184 hemi-pelvises) were included in the meta-analysis. The overall prevalence of the corona mortis in hemi-pelvises is high (49.3%). A venous corona mortis is more prevalent than an arterial corona mortis (41.7% vs. 17.0%). The corona mortis is more common in Asia (59.3%) than in Europe (42.8%) and North America (44.3%). Conclusions: As a corona mortis is present in an about half of all hemi-pelvises, it is important to consider the possibilities of its presence when undertaking surgical procedures and plan accordingly to avoid injuries. All surgeons operating in the retropubic region should have a thorough understanding of the anatomical characteristics and surgical implications of a corona mortis. © 2017 Published by Elsevier Ltd.
Introduction The definition of corona mortis is heterogeneous in the literature. This variability has caused discrepancies in the data of studies investigating this collateral circulation1. Darmanis et. al describes a corona mortis as a vascular connection between the obturator and external iliac or inferior epigastric arteries or veins located behind the superior pubic ramus in the retropubic space (space of Retzius) [1] (Fig. 1). Communicating vessels that traverse the superior pubic ramus are variously referred to as aberrant, anomalous, communicating,
* Corresponding author at: Department of Anatomy, Jagiellonian University Medical College, 12 Kopernika Street, 31-034, Krakow, Poland. E-mail address: [email protected] (B.M. Henry).
or variant vessels, or pubic branches of the obturator or inferior epigastric vessels [2]. However, not all of these vessels connect the obturator artery to the external iliac system, and thus, not all constitute a corona mortis vessel. Therefore, we define corona mortis as any abnormal anastomotic vessels between the external iliac and the obturator systems, excluding aberrant obturator arteries, as these originate from the external iliac artery or inferior epigastric artery and pierce the obturator membrane, not participating in the anastomosis [1]. Corona mortis varies extensively in terms of whether it is an arterial or venous connection, or both, as well as its laterality, side, and distance to pubic symphysis. The reported prevalence rates of venous and arterial types of corona mortis is highly variable, however, the venous type has consistently been reported to be more common than the arterial type [3–10]. Moreover, the corona mortis is located at a variable distance from the pubic symphysis,
https://doi.org/10.1016/j.injury.2017.12.007 0020-1383/© 2017 Published by Elsevier Ltd.
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
G Model JINJ 7527 No. of Pages 7
2
B. Sanna et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Fig. 1. Corona Mortis with both arterial and venous anastomosis.
with reported means ranging from 21 mm [7] to 90 mm [10]. Due to the significant variation in this collateral circulation, detailed anatomical knowledge of corona mortis is vital to enhance the prevention of possible iatrogenic errors in pubic region surgical procedures. The term corona mortis, or crown of death, reflects the clinical significance of this anatomical variant. Laceration of this collateral circulation may result in catastrophic bleeding, as the anastomotic vessels connect two high-volume systems that may retract into the obturator canal [2]. It has been reported to be a potential risk factor for significant hemorrhage in pelvic fractures [11–14], pelvic and acetabular surgeries [1,15,16], TVT-Secur procedures [17,18], surgery for paravaginal defects [19], oncological pelvic dissections [20], and conventional and laparoscopic hernia repair [21]. Studying vascular variations in corona mortis is crucial because of its association with a high risk of severe hemorrhage in hernia repair and other surgical procedures. Accurate anatomical knowledge of corona mortis may aid in reducing the incidence of surgical complications and improving the outcome of pubic surgical procedures. The aim of this meta-analysis was to systematically analyze and provide comprehensive data on the prevalence, anatomical characteristics, and ethnic variations of the corona mortis.
references of all included articles were searched to identify any further relevant articles. No date limits or language restrictions were applied. The authors strictly followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Supplement 1) for the entirety of the meta-analysis [22]. Study selection Study eligibility for inclusion in the meta-analysis was independently assessed by two separate reviewers. Studies were considered eligible for inclusion if they (1) provided clear anatomical definitions of corona mortis, (2) reported prevalence data concerning the corona mortis as primary or secondary outcome, and (3) were a cadaveric, imaging, or intraoperative study reporting extractable data. The following exclusion criteria were used: Case reports, editorials, conference abstracts, and studies reporting incomplete or irrelevant data. Articles in languages other than those spoken fluently by the authors were translated by medical professionals fluent in both the language of the original article, and English. All differences of opinion among the reviewers concerning the eligibility of the studies were solved by consensus through consultation with the author of the respective study.
Methods Data extraction Search strategy An extensive search was conducted through: PubMed, CNKI, Embase, ScienceDirect, Web of Science, SciELO, and BIOSIS, to identify articles eligible for inclusion in our meta-analysis. The search terms used were: corona mortis, coronae mortis, retropubic vascular anastomosis, retropubic vascular communication, obturator artery communication, and obturator artery anastomosis. The
Data from the included studies was independently extracted by two reviewers. Data extracted included investigative method, prevalence of the corona mortis among hemi-pelvises (overall, arterial, venous, and combined), distance from the corona mortis to pubic symphysis, gender, side, laterality, and ethnicity. In the event of disagreement among the authors, consultations with all the authors were conducted by consensus. If discrepancies in the
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
G Model JINJ 7527 No. of Pages 7
B. Sanna et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
data occurred, authors of the included studies were contacted for clarification and further data was sought after.
3
Results Study identification and characteristics of included studies
Statistical analysis The extracted data was pooled into a meta-analysis with a random-effects model using MetaXL version 5.0 (EpiGear International). The primary outcomes of the analysis were pooled prevalence estimate (PPE) of corona mortis. Secondary outcomes included PPE of side, laterality, distance from the corona mortis to pubic symphysis, arterial corona mortis (no venous corona mortis), venous corona mortis (no arterial corona mortis), and combined arterial and venous corona mortis. The Chi [2] test and the Higgins I [2] statistics were used to assess heterogeneity between the included studies and interpreted according to the guidelines in Chapter 9.5.2 of the Cochrane Handbook [23]. Subgroup analysis based on geographical distribution, modality of the study (cadaveric or intraoperative), gender, and side were conducted to detect sources of heterogeneity. To assess statistically significant differences between two or more subgroups, confidence intervals were compared. If the confidence intervals overlapped, the differences were considered statistically insignificant [24]. In addition, a leave-one-out sensitivity analysis was performed, as appropriate, to help further probe heterogeneity between the studies.
The study identification process is presented in Fig. 2. Major databases were extensively searched to initially identify 215 articles. A further 21 articles were identified through reference searching. One hundred and one articles were assessed by full text for potential eligibility. Of those, 80 articles were deemed ineligible. A total of 21 studies (N = 2184 hemi-pelvises) were considered eligible and included in the meta-analysis [1,2,6– 8,10,11,15,20,25–36]. There were 17 cadaveric studies [1,2,6– 8,10,15,25–27,29–32,34–36], 3 intraoperative studies [11,20,28], and 1 radiological study [33]). The studies spanned from 1996 to 2014 and displayed wide geographical distribution with studies hailing from Asia [6,26,27,29–31], Europe [1,3,7,8,11,15,20,28,32,34], North America [2,10,25,33], and South America35. The characteristics of the included studies are summarized in Table 1. Prevalence of the corona mortis Seventeen studies (N = 1664 hemi-pelvises) reported data on the prevalence of the corona mortis (Table 2) (Fig. 3). The total PPE of the corona mortis in the population was 49.3% (95%CI:37.2-61.5). In a subgroup analysis with respect to gender, males were found to have a higher PPE of the corona mortis (51.5%) (95%CI:31.1-71.7) compared to females (39.2%) (95%CI:21.7-58.2), albeit not significantly (Table 2). As for side, left-sided hemi-pelvises had a PPE of the corona mortis of 47.6% (95%CI:24.7-71.0), while right-sided hemi-pelvises had a PPE of 52.1% (95%CI:27.3-76.4) (Table 2). When analyzing the study modalities, cadaveric studies had a higher PPE (52.1%) (95%CI:37.5-66.5) than intraoperative studies (41.0%) (95% CI:14.9-69.7) (Table 2). In terms of geographical distribution, corona mortis was present in 59.3% (95%CI:29.2-86.5) in Asia, 44.3% (95%CI:30.3-58.6) in North America, and 42.8% (95%CI:28.258.1) in Europe (Table 2). No significant differences were noted when conducting a leave-one out analysis. Prevalence of an arterial corona mortis Thirteen studies (n = 1284 hemi-pelvises) reported data on the prevalence of an arterial only corona mortis (i.e. no venous corona mortis present) (Table 3). Our meta-analysis revealed an overall PPE of an arterial only corona mortis of 17.0% (95% CI 10.1-25.2). In a subgroup analysis of sides, left-sided hemi-pelvises had an arterial corona mortis in 8.8% of cases (95% CI 0.0-26.5), while in right-sided hemi-pelvises, arterial corona mortis was found in 12.4% of cases (95% CI 1.4-30.1) (Table 3). With respect to geographical distribution, North American studies had a PPE of 22.8% (95% CI 15.8-30.6), European studies had a PPE of 15.4% (95% CI:5.5-28.7), and Asian studies a PPE of 5.2% (95%CI:2.6-8.6) (Table 3). No significant differences were found in our leave-one out analysis. Prevalence of a venous corona mortis
Fig. 2. Flow diagram of included studies.
Thirteen studies (n = 850 hemi-pelvises) reported data on the prevalence of venous only corona mortis (no arterial corona mortis present) (Table 4). Overall, the PPE of a venous only corona mortis was 41.7% (95%CI:30.5-53.3). Subgroup analysis of sides revealed a left-sided PPE of a venous only corona mortis of 29.7% (95%CI:3.2-65.2), and a right-sided PPE of 25.3% (95%CI:0.0-76.0) (Table 4). In terms of a geographical analysis, Asian studies had a PPE of a venous only corona mortis of 43.0% (95%CI:28.3-58.4), European studies of 39.4% (95%CI:19.4-
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
G Model JINJ 7527 No. of Pages 7
4
B. Sanna et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Table 1 Table of included studies. Study ID
Country
Type of study
n of hemipelvises)
Reported CMOR prevalence (%)
Ates 2015 Baena 2015
Turkey Colombia
Intraoperative Cadaveric
398 28
Berberoglu 2001 Bible 2014 Darmanis 2007
Turkey USA England
Cadaveric Cadaveric Cadaveric
14 20 80
De Kleuver 1998 Drewes 2005 Hong 2004 Kacira 2011 Karakurt 2002 Kawai 2008 Mahato 2009 Namking 2006 Okcu 2004 Pellegrino 2014 Pungpapong 2005 Rusu 2010 Sarikcioglu 2003 Smith 2009 Stavropoulou-Deli 2013 Tornetta 1996
The Netherlands USA China Turkey Turkey Japan India Thailand Turkey Italy Thailand Romania Turkey USA Greece USA
Cadaveric Cadaveric Cadaveric Cadaveric Intraoperative Cadaveric Cadaveric Cadaveric Cadaveric Intraoperative Cadaveric Cadaveric Cadaveric Imaging Cadaveric Cadaveric
12 30 50 10 98 560 50 204 150 50 66 40 54 100 70 100
28.4 (arterial only) 35.7 (arterial only) 82.1 (venous only) 92.9 (venous only) 70.0 37.5 (arterial only) 66.3 (venous only) 50.0 46.7 72.0 40.0 28.6 21.6 62.0 76.0 60.7 56.0 77.3 65.0 20.4 29.0 25.7 42.0
Table 2 Total pooled prevalence estimates of a corona mortis in the population. Category
# of studies (# of hemi-pelvises)
Pooled Prevalence Estimate: % (95 CI)
I2: % (95 CI)1
Overall Males Females Left-sided Right-sided Cadaveric studies Intraoperative studies Asia Europe North America
17 (1664) 3 (219) 4 (149) 5 (241) 5 (233) 14 (1416) 2 (148) 6 (940) 7 (474) 4 (250)
49.3 (37.2–61.5) 51.5 (31.1–71.7) 39.2 (21.7–58.2) 47.6 (24.7–71.0) 52.1 (27.3–76.4) 52.1 (37.5–66.5) 41.0 (14.9–69.7) 59.3 (29.2–86.5) 42.8 (28.2–58.1) 44.3 (30.3–58.6)
95.3 (93.8–96.5) 88.0 (66.5–95.7) 77.0 (37.3–91.5) 91.6 (83.3–95.7) 92.1 (84.7–95.9) 96.0 (94.6–97.0) 90.3 (64.7–94.3) 98.1 (97.3–98.7) 90.0 (82.1–94.5) 77.1 (37.8–91.6)
1
For Cochran’s Q, there was a statistically significant amount of heterogeneity between studies (p < 0.10) for all analyses.
Fig. 3. Forest plot of the prevalence of the Corona Mortis.
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
G Model JINJ 7527 No. of Pages 7
B. Sanna et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
5
Table 3 Pooled prevalence estimates of an arterial only corona mortis. Category
# of studies (# of hemi-pelvises)
Pooled Prevalence Estimate: % (95 CI)
I2: % (95 CI)
Cochran’s Q p-values
Overall Left Right Asia Europe North America
13 (1284) 3 (188) 3 (186) 2 (214) 6 (792) 4 (250)
17.0 (10.1–25.2) 8.8 (0.0 26.5) 12.4 (1.4–30.1) 5.2 (2.6–8.6) 15.4 (5.5–28.7) 22.8 (15.8–30.6)
91.0 (86.5–94.0) 88.2 (67.2–95.7) 86.5 (61.2–95.3) 0.0 (0.0 0.0) 93.8 (89.2–96.5) 42.5 (0.0 80.6)
<0.001 <0.001 0.001 0.587 <0.001 0.156
Table 4 Pooled prevalence estimates of a venous only corona mortis. Category
# of studies (# of hemi-pelvises)
Pooled Prevalence Estimate: % (95 CI)
I2: % (95 CI)
Cochran’s Q p-values
Overall Left Right Asia Europe North America
13 (850) 3 (166) 3 (158) 3 (264) 6 (408) 3 (150)
41.7 (30.5–53.3) 29.7 (3.2–65.2) 25.3 (0.0 76.0) 43.0 (28.3–58.4) 39.4 (19.4–61.3) 34.8 (27.4–32.6)
90.4 (84.5–93.6) 93.4 (84.1–97.2) 96.6 (93.1–98.3) 58.9 (0.0 90.9) 94.1 (89.8–96.6) 0.0 (0.0 60.4)
<0.001 <0.001 <0.001 0.040 <0.001 0.769
Table 5 Pooled prevalence estimates of a combined arterial and venous corona mortis. Category
# of studies (# of hemi-pelvises)
Pooled Prevalence Estimate: % (95 CI)
I2: % (95 CI)
Cochran’s Q p-values
Overall Asia Europe
6 (508) 3 (264) 3 (244)
15.2 (6.2–26.9) 18.4 (13.9–23.3) 12.5 (0.0 35.8)
88.2 (76.9–94.0) 0.0 (0.0 74.2) 94.3 (87.0 97.5)
<0.001 0.668 <0.001
61.3), and North American studies of 34.8% (95%CI:27.4-32.6) (Table 4). No significant differences were noted in a leave-one out analysis. Prevalence of a combined arterial and venous corona mortis Six studies (n = 508 hemi-pelvises) reported data on the prevalence of a combined arterial and venous corona mortis (Table 5). The overall PPE of a combined arterial and venous corona mortis was 15.2% (95%CI:6.2-26.9). In terms of geographical distribution, Asian studies reported a PPE of a combined arterial and venous corona mortis of 18.4% (95% CI:13.9-23.3), while European studies reported a PPE of 12.5% (95% CI:0.0-35.8) (Table 5). No significant differences were noted when conducting a leave-one out analysis. Laterality of the corona mortis Six studies (n = 240 patients) reported data on the prevalence of bilaterally present corona mortis’. Overall, the PPE of bilateral corona mortis’ was 17.9% (95%CI:9.3-28.4). No significant differences were noted in our leave-one out analysis. Distance from the corona mortis to pubic symphysis Three studies (n = 67 arterial corona mortis) reported data on the distance from the corona mortis to pubic symphysis in arterial corona mortis’. The pooled mean length of the distance from the corona mortis to pubic symphysis in arterial corona mortis’ was 59.90 mm (95%CI:49.05-70.75). Three studies (n = 76 venous corona mortis) reported data on the distance from the corona mortis to pubic symphysis in venous corona mortis’. Our meta-analysis revealed that the pooled mean length from the corona mortis to pubic symphysis in venous corona mortis’ was 50.70 mm (95% CI: 35.97-65.42).
Discussion Accidental transection of the corona mortis in pubic surgical procedures can result in a potentially life-threatening hemorrhage. The high incidence of this vascular connection in the population makes it particularly clinically significant. Detailed anatomical knowledge of its prevalence, variable anastomotic connections, and morphological characteristics is therefore vital to reduce the risks of possible iatrogenic errors in hernia repair and other surgical procedures in the pubic region. The aim of our study was to establish the true prevalence of the corona mortis, its associated anatomical characteristics, and variation between ethnic groups, by pooling together all available studies through a meta-analysis. The presence of corona mortis is a risk factor for complications in pubic rami fractures and in several surgical procedures [17,19– 21,37]. The coexistence of vascular variations originating from either the external iliac, inferior epigastric, or femoral vessels have been reported to require great attention from surgeons during surgical repair of inguinal and femoral hernias [11,21,32]. When present, the corona mortis can be damaged during fixation of the mesh on Cooper’s ligament in a laparoscopic herniorrhaphy, in which a staple or a tack is driven blindly into Cooper’s ligament [32]. This can cause uncontrollable bleeding, retroperitoneal hematoma, conversion to open surgery and reoperation [11]. During conventional hernia repair, damage to the anastomosis may result in significant hemorrhage, but this can usually be controlled by extension of the exposure [38]. An exposed corona mortis during oncological pelvic dissections also emphasizes the importance of exercising extreme caution when dissecting malignancies close to this collateral circulation [20]. However, even though corona mortis vessels were identified behind the superior pubic ramus in anatomical dissections during lymphadenectomies [20], another study reported that these corona mortis vessels did not seem to be as great a threat in clinical practice during these procedures [1].
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
G Model JINJ 7527 No. of Pages 7
6
B. Sanna et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Other serious complications reported include injury to the corona mortis following a TVT-Secur procedure with a U-approach, in which a tape is pushed behind the arc of the pubic bone, and the contact between the sharp, scalpel-like incisor and the vessel could cause damage to the corona mortis [39] resulting in severe bleeding into the space of Retzius [17]. Moreover, as the corona mortis is located behind the superior pubic ramus, it is susceptible to injury in scalpel dissections [37]. In addition, extreme care should be taken in the case of injury to the tendinous arch of pelvic fascia during paravaginal repair due to its close relation with the corona mortis [19]. To avoid iatrogenic errors in the abovementioned surgical procedures, radiologists should alert surgeons regarding the digital anatomical data of the corona mortis and illustrate the appropriate operation approach, reducing the risk of lacerating corona mortis vessels [37,40]. Our main findings revealed that the overall prevalence of the corona mortis in the population is high (49.3%), that it is more common in Asia (59.3%) than in Europe (42.8%) and North America (44.3%), and that a venous corona mortis (41.7%) is more commonly encountered than an arterial corona mortis (17.0%). In terms of geographical subgroups, the statistically significant difference between Asia, and Europe and North America, can be explained due to a more similar racial composition of the population and genetic similarities between Europeans and North Americans, as compared to Asians. Another explanation for the discrepancy in these findings could be the confusion in the literature over the term corona mortis and what it should include. The definition of the corona mortis used in our paper is the widely accepted definition [1,3] and should be consequently used in all populations to avoid unnecessary discrepancies in studies investing the corona mortis. Minimally invasive procedures, such as the insertion of the TVTSecur, have been linked to injury of the corona mortis due to the vessel’s position close to the place of the margin (25–50 mm from the pubic symphysis) [34]. The degree of individual variation concerning the position of the corona mortis, and the fact that both an arterial and venous corona mortis may be present, and in 17.9% of cases even bilaterally, highlights that a direct contact with the vessel is more probable in cases in which corona mortis vessels are located close to pubic symphysis. Moreover, surgeons should bear in mind that although the damage to an arterial corona mortis is more dangerous in terms of hemorrhage and hemodynamic instability, the identification of an active venous source of bleeding is more difficult and angiography cannot be used for this purpose [34]. A heightened awareness of the possibility of contact with the corona mortis should help reduce hemorrhage rates during minimally invasive procedures. Our meta-analysis was limited by the high amount of heterogeneity between studies. This can be explained by regional differences in the development of the vascular system, difficulties in dissecting the collapsed vessels in cadavers, and the limitations caused by the size of the vessels [7]. Small vessels often may be excluded, however, vessels smaller than 2 mm in diameter are especially important to map as they are poorly visible and harder to protect than larger vessels, increasing the risk of hemorrhagic complications in patients with small corona mortis vessels. There are few intraoperative studies investigating the corona mortis available in the literature. To be able to properly compare cadaveric and intraoperative studies, more intraoperative studies are needed. However, the large amount of cadaveric studies and their large sample sizes is a strength to our meta-analysis, as these studies are the gold standard for this type of anatomical investigation. In addition, future studies should be conducted in continents poorly represented by our meta-analysis, such as North America, South America, Africa, and Oceania to give a better
estimate of the true PPE of the corona mortis in various populations. Considering its clinical significance in numerous surgical procedures, the corona mortis, or Crown of Death, is an important anatomic variation that should be treated with the same respect as its name implies. Conclusions Our main findings revealed that the corona mortis was found to be present in nearly half of the general population, that a venous corona mortis is 2–3 x more prevalent than an arterial corona mortis, and that the corona mortis is more common in Asia than in Europe and North America. Based on our anatomical findings, we recommend an effort should be taken to identify and preserve the anastomotic vessels to prevent serious iatrogenic errors in surgical procedures in the retropubic region.
Acknowledgements Krzysztof A. Tomaszewski was supported by the Foundation for Polish Science (FNP) and the Polish Ministry of Science and Higher Education Grant for Young Scientists. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.injury.2017.12.007. References [1] Darmanis S, Lewis A, Mansoor A, Bircher M. Corona mortis: an anatomical study with clinical implications in approaches to the pelvis and acetabulum. Clin Anat 2007;20:433–9. [2] Drewes PG, Marinis SI, Schaffer JI, Boreham MK, Corton MM. Vascular anatomy over the superior pubic rami in female cadavers. Am J Obstet Gynecol 2005;193:2165–8. lu M, Uz A, Ozmen MM, Bozkurt MC, Erkuran C, Taner S, et al. Corona [3] Berberog mortis: an anatomic study in seven cadavers and an endoscopic study in 28 patients. Surg Endosc 2001;15:72–5. [4] Gilroy AM, Hermey DC, DiBenedetto LM, Marks Jr SC, Page DW, Lei QF. Variability of the obturator vessels. Clin Anat 1997;10:328–32. [5] Jakubowicz M, Czarniawska-Grzesiñska M. Variability in origin and topography of the inferior epigastric and obturator arteries. Folia Morphol (Warsz) 1996;55:121–6. [6] Namking M, Woraputtaporn W, Buranarugsa M, Kerdkoonchorn M. Corona mortis in thai. FASEB Journal 2006;20:A445. [7] Okcu G, Erkan S, Yercan HS, Ozic U. The incidence and location of corona mortis: a study on 75 cadavers. Acta Orthop Scand 2004;75:53–5. [8] Sarikcioglu L, Sindel M, Akyildiz F, Gur S. Anastomotic vessels in the retropubic region: corona mortis. Folia Morphol (Warsz) 2003;62:179–82. [9] Teague DC, Graney DO, Routt ML. Retropubic vascular hazards of the ilioinguinal exposure: a cadaveric and clinical study. J Orthop Trauma 1996;10:156–9. [10] Tornetta P, Hochwald N, Levine R. Corona mortis. Incidence and location. Clin Orthop Relat Res 1996;329:97–101. [11] Ates M, Kinaci E, Kose E, Soyer V, Sarici B, Cuglan S, et al. Corona mortis: in vivo anatomical knowledge and the risk of injury in totally extraperitoneal inguinal hernia repair. Hernia 2016;20:659–65. [12] Daeubler B, Anderson SE, Leunig M, Triller J. Hemorrhage secondary to pelvic fracture: coil embolization of an aberrant obturator artery. J Endovasc Ther 2003;10:676–80. [13] Garrido-Gómez J, Pena-Rodríguez C, Martín-Noguerol T, Hernández-Cortes P. Corona mortis artery avulsion due to a stable pubic ramus fracture. Orthopedics 2012;35:e80–82. [14] Henning P, Brenner B, Brunner K, Zimmermann H. Hemodynamic instability following an avulsion of the corona mortis artery secondary to a benign pubic ramus fracture. J Trauma Inj Infect Crit Care 2007;62:E14–7. [15] de Kleuver M, Kooijman MA, Kauer JM, Veth RP. Pelvic osteotomies: anatomic pitfalls at the pubic bone. A cadaver study. Arch Orthop Trauma Surg 1998;117:270–2. [16] Damsin JP, Lazennec JY, Gonzales M, Guerin-Surville H, Hannoun L. Arterial supply of the acetabulum in the fetus: application to periacetabular surgery in childhood. Surg Radiol Anat 1992;14:215–21.
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
G Model JINJ 7527 No. of Pages 7
B. Sanna et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx [17] Larsson P-G, Teleman P, Persson J. A serious bleeding complication with injury of the corona mortis with the TVT-Secur procedure. Int Urogynecol J 2010;21:1175–7. [18] Gobrecht U, Kuhn A, Fellman B. Injury of the corona mortis during vaginal tape insertion (TVT-SecurTM using the U-Approach). Int Urogynecol J 2011;22:443– 5. [19] Ersoy M, Sagsoz N, Bozkurt MC, Apaydin N, Elhan A, Tekdemir I. Important anatomical structures used in paravaginal defect repair: cadaveric study. Eur J Obstet Gynecol Reprod Biol 2004;112:206–13. [20] Pellegrino A, Damiani GR, Marco S, Ciro S, Cofelice V, Rosati F. Corona mortis exposition during laparoscopic procedure for gynecological malignancies. Updates Surg 2014;66:65–8. [21] Ramser M, Messmer AS, Zbinden I, et al. Incarcerated obturator hernialaparoscopic repair with intraoperative view of the corona mortis. J Surg Case Reports 2014;2014(8):rju081. [22] Moher D, Liberati A, Tetzlaff J, Altman D. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009;6(7):1000097. [23] Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions Cochrane Book Series. THE COCHRANE COLLABORATION 1; 2011. [24] Henry BM, Tomaszewski KA, Walocha JA. Methods of Evidence-Based Anatomy: a guide to conducting systematic reviews and meta-analysis of anatomical studies. Ann Anat 2016;205:16–21. [25] Bible JE, Choxi AA, Kadakia RJ, Evans JM, Mir HR. Quantification of bony pelvic exposure through the modified Stoppa approach. J Orthop Trauma 2014;28:320–3. [26] Hong H, Pan Z, Chen X, Huang Z. An anatomical study of corona mortis and its clinical significance. Chin J Traumatol: Zhonghua chuang shang za zhi 2004;7:165–9. [27] Kacra BK, Arazi M, Cicekcibasi AE, Buyukmumcu M, Demirci S. Modified medial stoppa approach for acetabular fractures: an anatomic study. J Trauma Inj Infect Crit Care 2011;71:1340–4.
7
[28] Karakurt L, Karaca I, Yilmaz E, Burma O, Serin E. Corona mortis: incidence and location. Arch Orthop Trauma Surg 2002;122:163–4. [29] Kawai K, Honma S, Koizumi M, Kodama K. Inferior epigastric artery arising from the obturator artery as a terminal branch of the internal iliac artery and consideration of its rare occurrence. Ann Anat 2008;190:541–8. [30] Mahato NK. Retro-pubic vascular anomalies: a study of abnormal obturator vessels. Eur J Anat 2009;13:121–6. [31] Pungpapong S, Thum-umnauysuk S. Incidence of corona mortis; preperitoneal anatomy for laparoscopic hernia repair. J Med Assoc Thai 2005;88(Suppl. 4): S51–3. [32] Rusu MC, Cergan R, Motoc AG, Folescu R, Pop E. Anatomical considerations on the corona mortis. Surg Radiol Anat 2010;32:17–24. [33] Smith JC, Gregorius JC, Breazeale BH, Watkins GE. The corona mortis, a frequent vascular variant susceptible to blunt pelvic trauma: identification at routine multidetector CT. J Vasc Interv Radiol 2009;20:455–60. [34] Stavropoulou-Deli A, Anagnostopoulou S. Corona mortis: anatomical data and clinical considerations. Aust New Zeal J Obstet Gynaecol 2013;53:283–6. [35] Baena G, Rojas S, Peña E. Corona mortis: anatomical and clinical relevance and occurrence in a sample of the colombian population. Int J Morphol 2015;33:130–6. lu M, Uz A, Ozmen MM, Bozkurt MC, Erkuran C, Taner S, et al. Corona [36] Berberog mortis: an anatomic study in seven cadavers and an endoscopic study in 28 patients. Surg Endosc 2001;15:72–5. [37] Al Talalwah W. A new concept and classification of corona mortis and its clinical significance. Chin J Traumatol: Zhonghua chuang shang za zhi 2016;19:251–4. [38] Gaines RD. Complications of groin hernia repair: their prevention and management. J Natl Med Assoc 1978;70:195–8. [39] Castellani D, Saldutto P, Galica V, Masciovecchio S, Galatioto GP, Vicentini C. Transobturator surgery in presence of corona mortis: a study in 13 women. Urol 2016;83:200–3. [40] Han Y, Liu P, Chen C, Duan H, Chen L, Xu Y, et al. A digital anatomical study of the corona mortis in females. Minim Invasive Ther Allied Technol 2016;26:111–8.
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery Beatrice Sannaa,b , Brandon Michael Henrya,c,* , Jens Viksea,d , Bendik Skinningsruda,c , Jakub R. Pe˛kalaa,c, Jerzy A. Walochaa,c, Roberto Cirocchie , Krzysztof A. Tomaszewskia,c a
International Evidence-Based Anatomy Working Group, Krakow, Poland Faculty of Medicine & Surgery, University of Cagliari, Sardinia, Italy Department of Anatomy, Jagiellonian University Medical College, Krakow, Poland d Division of Medicine, Stavanger University Hospital, Stavanger, Norway e Department of Surgical Sciences, University of Perugia, Italy b c
A R T I C L E I N F O
Keywords: Corona mortis Obturator vessels Hernia repair Pelvic lymphadenectomy Retropubic region Pelvic fracture Vascular anatomy
A B S T R A C T
Purpose: Corona mortis is a highly variable vascular connection between the obturator and external iliac or inferior epigastric arteries or veins located behind the superior pubic ramus in the retropubic space (space of Retzius). Due to the significant variation in this collateral circulation, detailed anatomical knowledge of the corona mortis is vital to enhance the prevention of possible iatrogenic errors in hernia repair and other pubic surgical procedures. The aim of our meta-analysis was to provide comprehensive data on the prevalence, anatomical characteristics, and ethnic variations of the corona mortis vessel. Methods: An extensive search was conducted through the major electronic databases to identify eligible articles. Data extracted included investigative method, prevalence of the corona mortis vessels among hemi-pelvises (overall, arterial only, venous only, and combined), distance from the corona mortis to pubic symphysis, and assessment of gender, side, laterality, and ethnicity subgroups. Results: A total of 21 studies (n = 2184 hemi-pelvises) were included in the meta-analysis. The overall prevalence of the corona mortis in hemi-pelvises is high (49.3%). A venous corona mortis is more prevalent than an arterial corona mortis (41.7% vs. 17.0%). The corona mortis is more common in Asia (59.3%) than in Europe (42.8%) and North America (44.3%). Conclusions: As a corona mortis is present in an about half of all hemi-pelvises, it is important to consider the possibilities of its presence when undertaking surgical procedures and plan accordingly to avoid injuries. All surgeons operating in the retropubic region should have a thorough understanding of the anatomical characteristics and surgical implications of a corona mortis. © 2017 Published by Elsevier Ltd.
Introduction The definition of corona mortis is heterogeneous in the literature. This variability has caused discrepancies in the data of studies investigating this collateral circulation1. Darmanis et. al describes a corona mortis as a vascular connection between the obturator and external iliac or inferior epigastric arteries or veins located behind the superior pubic ramus in the retropubic space (space of Retzius) [1] (Fig. 1). Communicating vessels that traverse the superior pubic ramus are variously referred to as aberrant, anomalous, communicating,
* Corresponding author at: Department of Anatomy, Jagiellonian University Medical College, 12 Kopernika Street, 31-034, Krakow, Poland. E-mail address: [email protected] (B.M. Henry).
or variant vessels, or pubic branches of the obturator or inferior epigastric vessels [2]. However, not all of these vessels connect the obturator artery to the external iliac system, and thus, not all constitute a corona mortis vessel. Therefore, we define corona mortis as any abnormal anastomotic vessels between the external iliac and the obturator systems, excluding aberrant obturator arteries, as these originate from the external iliac artery or inferior epigastric artery and pierce the obturator membrane, not participating in the anastomosis [1]. Corona mortis varies extensively in terms of whether it is an arterial or venous connection, or both, as well as its laterality, side, and distance to pubic symphysis. The reported prevalence rates of venous and arterial types of corona mortis is highly variable, however, the venous type has consistently been reported to be more common than the arterial type [3–10]. Moreover, the corona mortis is located at a variable distance from the pubic symphysis,
https://doi.org/10.1016/j.injury.2017.12.007 0020-1383/© 2017 Published by Elsevier Ltd.
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
G Model JINJ 7527 No. of Pages 7
2
B. Sanna et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Fig. 1. Corona Mortis with both arterial and venous anastomosis.
with reported means ranging from 21 mm [7] to 90 mm [10]. Due to the significant variation in this collateral circulation, detailed anatomical knowledge of corona mortis is vital to enhance the prevention of possible iatrogenic errors in pubic region surgical procedures. The term corona mortis, or crown of death, reflects the clinical significance of this anatomical variant. Laceration of this collateral circulation may result in catastrophic bleeding, as the anastomotic vessels connect two high-volume systems that may retract into the obturator canal [2]. It has been reported to be a potential risk factor for significant hemorrhage in pelvic fractures [11–14], pelvic and acetabular surgeries [1,15,16], TVT-Secur procedures [17,18], surgery for paravaginal defects [19], oncological pelvic dissections [20], and conventional and laparoscopic hernia repair [21]. Studying vascular variations in corona mortis is crucial because of its association with a high risk of severe hemorrhage in hernia repair and other surgical procedures. Accurate anatomical knowledge of corona mortis may aid in reducing the incidence of surgical complications and improving the outcome of pubic surgical procedures. The aim of this meta-analysis was to systematically analyze and provide comprehensive data on the prevalence, anatomical characteristics, and ethnic variations of the corona mortis.
references of all included articles were searched to identify any further relevant articles. No date limits or language restrictions were applied. The authors strictly followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Supplement 1) for the entirety of the meta-analysis [22]. Study selection Study eligibility for inclusion in the meta-analysis was independently assessed by two separate reviewers. Studies were considered eligible for inclusion if they (1) provided clear anatomical definitions of corona mortis, (2) reported prevalence data concerning the corona mortis as primary or secondary outcome, and (3) were a cadaveric, imaging, or intraoperative study reporting extractable data. The following exclusion criteria were used: Case reports, editorials, conference abstracts, and studies reporting incomplete or irrelevant data. Articles in languages other than those spoken fluently by the authors were translated by medical professionals fluent in both the language of the original article, and English. All differences of opinion among the reviewers concerning the eligibility of the studies were solved by consensus through consultation with the author of the respective study.
Methods Data extraction Search strategy An extensive search was conducted through: PubMed, CNKI, Embase, ScienceDirect, Web of Science, SciELO, and BIOSIS, to identify articles eligible for inclusion in our meta-analysis. The search terms used were: corona mortis, coronae mortis, retropubic vascular anastomosis, retropubic vascular communication, obturator artery communication, and obturator artery anastomosis. The
Data from the included studies was independently extracted by two reviewers. Data extracted included investigative method, prevalence of the corona mortis among hemi-pelvises (overall, arterial, venous, and combined), distance from the corona mortis to pubic symphysis, gender, side, laterality, and ethnicity. In the event of disagreement among the authors, consultations with all the authors were conducted by consensus. If discrepancies in the
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
G Model JINJ 7527 No. of Pages 7
B. Sanna et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
data occurred, authors of the included studies were contacted for clarification and further data was sought after.
3
Results Study identification and characteristics of included studies
Statistical analysis The extracted data was pooled into a meta-analysis with a random-effects model using MetaXL version 5.0 (EpiGear International). The primary outcomes of the analysis were pooled prevalence estimate (PPE) of corona mortis. Secondary outcomes included PPE of side, laterality, distance from the corona mortis to pubic symphysis, arterial corona mortis (no venous corona mortis), venous corona mortis (no arterial corona mortis), and combined arterial and venous corona mortis. The Chi [2] test and the Higgins I [2] statistics were used to assess heterogeneity between the included studies and interpreted according to the guidelines in Chapter 9.5.2 of the Cochrane Handbook [23]. Subgroup analysis based on geographical distribution, modality of the study (cadaveric or intraoperative), gender, and side were conducted to detect sources of heterogeneity. To assess statistically significant differences between two or more subgroups, confidence intervals were compared. If the confidence intervals overlapped, the differences were considered statistically insignificant [24]. In addition, a leave-one-out sensitivity analysis was performed, as appropriate, to help further probe heterogeneity between the studies.
The study identification process is presented in Fig. 2. Major databases were extensively searched to initially identify 215 articles. A further 21 articles were identified through reference searching. One hundred and one articles were assessed by full text for potential eligibility. Of those, 80 articles were deemed ineligible. A total of 21 studies (N = 2184 hemi-pelvises) were considered eligible and included in the meta-analysis [1,2,6– 8,10,11,15,20,25–36]. There were 17 cadaveric studies [1,2,6– 8,10,15,25–27,29–32,34–36], 3 intraoperative studies [11,20,28], and 1 radiological study [33]). The studies spanned from 1996 to 2014 and displayed wide geographical distribution with studies hailing from Asia [6,26,27,29–31], Europe [1,3,7,8,11,15,20,28,32,34], North America [2,10,25,33], and South America35. The characteristics of the included studies are summarized in Table 1. Prevalence of the corona mortis Seventeen studies (N = 1664 hemi-pelvises) reported data on the prevalence of the corona mortis (Table 2) (Fig. 3). The total PPE of the corona mortis in the population was 49.3% (95%CI:37.2-61.5). In a subgroup analysis with respect to gender, males were found to have a higher PPE of the corona mortis (51.5%) (95%CI:31.1-71.7) compared to females (39.2%) (95%CI:21.7-58.2), albeit not significantly (Table 2). As for side, left-sided hemi-pelvises had a PPE of the corona mortis of 47.6% (95%CI:24.7-71.0), while right-sided hemi-pelvises had a PPE of 52.1% (95%CI:27.3-76.4) (Table 2). When analyzing the study modalities, cadaveric studies had a higher PPE (52.1%) (95%CI:37.5-66.5) than intraoperative studies (41.0%) (95% CI:14.9-69.7) (Table 2). In terms of geographical distribution, corona mortis was present in 59.3% (95%CI:29.2-86.5) in Asia, 44.3% (95%CI:30.3-58.6) in North America, and 42.8% (95%CI:28.258.1) in Europe (Table 2). No significant differences were noted when conducting a leave-one out analysis. Prevalence of an arterial corona mortis Thirteen studies (n = 1284 hemi-pelvises) reported data on the prevalence of an arterial only corona mortis (i.e. no venous corona mortis present) (Table 3). Our meta-analysis revealed an overall PPE of an arterial only corona mortis of 17.0% (95% CI 10.1-25.2). In a subgroup analysis of sides, left-sided hemi-pelvises had an arterial corona mortis in 8.8% of cases (95% CI 0.0-26.5), while in right-sided hemi-pelvises, arterial corona mortis was found in 12.4% of cases (95% CI 1.4-30.1) (Table 3). With respect to geographical distribution, North American studies had a PPE of 22.8% (95% CI 15.8-30.6), European studies had a PPE of 15.4% (95% CI:5.5-28.7), and Asian studies a PPE of 5.2% (95%CI:2.6-8.6) (Table 3). No significant differences were found in our leave-one out analysis. Prevalence of a venous corona mortis
Fig. 2. Flow diagram of included studies.
Thirteen studies (n = 850 hemi-pelvises) reported data on the prevalence of venous only corona mortis (no arterial corona mortis present) (Table 4). Overall, the PPE of a venous only corona mortis was 41.7% (95%CI:30.5-53.3). Subgroup analysis of sides revealed a left-sided PPE of a venous only corona mortis of 29.7% (95%CI:3.2-65.2), and a right-sided PPE of 25.3% (95%CI:0.0-76.0) (Table 4). In terms of a geographical analysis, Asian studies had a PPE of a venous only corona mortis of 43.0% (95%CI:28.3-58.4), European studies of 39.4% (95%CI:19.4-
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
G Model JINJ 7527 No. of Pages 7
4
B. Sanna et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Table 1 Table of included studies. Study ID
Country
Type of study
n of hemipelvises)
Reported CMOR prevalence (%)
Ates 2015 Baena 2015
Turkey Colombia
Intraoperative Cadaveric
398 28
Berberoglu 2001 Bible 2014 Darmanis 2007
Turkey USA England
Cadaveric Cadaveric Cadaveric
14 20 80
De Kleuver 1998 Drewes 2005 Hong 2004 Kacira 2011 Karakurt 2002 Kawai 2008 Mahato 2009 Namking 2006 Okcu 2004 Pellegrino 2014 Pungpapong 2005 Rusu 2010 Sarikcioglu 2003 Smith 2009 Stavropoulou-Deli 2013 Tornetta 1996
The Netherlands USA China Turkey Turkey Japan India Thailand Turkey Italy Thailand Romania Turkey USA Greece USA
Cadaveric Cadaveric Cadaveric Cadaveric Intraoperative Cadaveric Cadaveric Cadaveric Cadaveric Intraoperative Cadaveric Cadaveric Cadaveric Imaging Cadaveric Cadaveric
12 30 50 10 98 560 50 204 150 50 66 40 54 100 70 100
28.4 (arterial only) 35.7 (arterial only) 82.1 (venous only) 92.9 (venous only) 70.0 37.5 (arterial only) 66.3 (venous only) 50.0 46.7 72.0 40.0 28.6 21.6 62.0 76.0 60.7 56.0 77.3 65.0 20.4 29.0 25.7 42.0
Table 2 Total pooled prevalence estimates of a corona mortis in the population. Category
# of studies (# of hemi-pelvises)
Pooled Prevalence Estimate: % (95 CI)
I2: % (95 CI)1
Overall Males Females Left-sided Right-sided Cadaveric studies Intraoperative studies Asia Europe North America
17 (1664) 3 (219) 4 (149) 5 (241) 5 (233) 14 (1416) 2 (148) 6 (940) 7 (474) 4 (250)
49.3 (37.2–61.5) 51.5 (31.1–71.7) 39.2 (21.7–58.2) 47.6 (24.7–71.0) 52.1 (27.3–76.4) 52.1 (37.5–66.5) 41.0 (14.9–69.7) 59.3 (29.2–86.5) 42.8 (28.2–58.1) 44.3 (30.3–58.6)
95.3 (93.8–96.5) 88.0 (66.5–95.7) 77.0 (37.3–91.5) 91.6 (83.3–95.7) 92.1 (84.7–95.9) 96.0 (94.6–97.0) 90.3 (64.7–94.3) 98.1 (97.3–98.7) 90.0 (82.1–94.5) 77.1 (37.8–91.6)
1
For Cochran’s Q, there was a statistically significant amount of heterogeneity between studies (p < 0.10) for all analyses.
Fig. 3. Forest plot of the prevalence of the Corona Mortis.
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
G Model JINJ 7527 No. of Pages 7
B. Sanna et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
5
Table 3 Pooled prevalence estimates of an arterial only corona mortis. Category
# of studies (# of hemi-pelvises)
Pooled Prevalence Estimate: % (95 CI)
I2: % (95 CI)
Cochran’s Q p-values
Overall Left Right Asia Europe North America
13 (1284) 3 (188) 3 (186) 2 (214) 6 (792) 4 (250)
17.0 (10.1–25.2) 8.8 (0.0 26.5) 12.4 (1.4–30.1) 5.2 (2.6–8.6) 15.4 (5.5–28.7) 22.8 (15.8–30.6)
91.0 (86.5–94.0) 88.2 (67.2–95.7) 86.5 (61.2–95.3) 0.0 (0.0 0.0) 93.8 (89.2–96.5) 42.5 (0.0 80.6)
<0.001 <0.001 0.001 0.587 <0.001 0.156
Table 4 Pooled prevalence estimates of a venous only corona mortis. Category
# of studies (# of hemi-pelvises)
Pooled Prevalence Estimate: % (95 CI)
I2: % (95 CI)
Cochran’s Q p-values
Overall Left Right Asia Europe North America
13 (850) 3 (166) 3 (158) 3 (264) 6 (408) 3 (150)
41.7 (30.5–53.3) 29.7 (3.2–65.2) 25.3 (0.0 76.0) 43.0 (28.3–58.4) 39.4 (19.4–61.3) 34.8 (27.4–32.6)
90.4 (84.5–93.6) 93.4 (84.1–97.2) 96.6 (93.1–98.3) 58.9 (0.0 90.9) 94.1 (89.8–96.6) 0.0 (0.0 60.4)
<0.001 <0.001 <0.001 0.040 <0.001 0.769
Table 5 Pooled prevalence estimates of a combined arterial and venous corona mortis. Category
# of studies (# of hemi-pelvises)
Pooled Prevalence Estimate: % (95 CI)
I2: % (95 CI)
Cochran’s Q p-values
Overall Asia Europe
6 (508) 3 (264) 3 (244)
15.2 (6.2–26.9) 18.4 (13.9–23.3) 12.5 (0.0 35.8)
88.2 (76.9–94.0) 0.0 (0.0 74.2) 94.3 (87.0 97.5)
<0.001 0.668 <0.001
61.3), and North American studies of 34.8% (95%CI:27.4-32.6) (Table 4). No significant differences were noted in a leave-one out analysis. Prevalence of a combined arterial and venous corona mortis Six studies (n = 508 hemi-pelvises) reported data on the prevalence of a combined arterial and venous corona mortis (Table 5). The overall PPE of a combined arterial and venous corona mortis was 15.2% (95%CI:6.2-26.9). In terms of geographical distribution, Asian studies reported a PPE of a combined arterial and venous corona mortis of 18.4% (95% CI:13.9-23.3), while European studies reported a PPE of 12.5% (95% CI:0.0-35.8) (Table 5). No significant differences were noted when conducting a leave-one out analysis. Laterality of the corona mortis Six studies (n = 240 patients) reported data on the prevalence of bilaterally present corona mortis’. Overall, the PPE of bilateral corona mortis’ was 17.9% (95%CI:9.3-28.4). No significant differences were noted in our leave-one out analysis. Distance from the corona mortis to pubic symphysis Three studies (n = 67 arterial corona mortis) reported data on the distance from the corona mortis to pubic symphysis in arterial corona mortis’. The pooled mean length of the distance from the corona mortis to pubic symphysis in arterial corona mortis’ was 59.90 mm (95%CI:49.05-70.75). Three studies (n = 76 venous corona mortis) reported data on the distance from the corona mortis to pubic symphysis in venous corona mortis’. Our meta-analysis revealed that the pooled mean length from the corona mortis to pubic symphysis in venous corona mortis’ was 50.70 mm (95% CI: 35.97-65.42).
Discussion Accidental transection of the corona mortis in pubic surgical procedures can result in a potentially life-threatening hemorrhage. The high incidence of this vascular connection in the population makes it particularly clinically significant. Detailed anatomical knowledge of its prevalence, variable anastomotic connections, and morphological characteristics is therefore vital to reduce the risks of possible iatrogenic errors in hernia repair and other surgical procedures in the pubic region. The aim of our study was to establish the true prevalence of the corona mortis, its associated anatomical characteristics, and variation between ethnic groups, by pooling together all available studies through a meta-analysis. The presence of corona mortis is a risk factor for complications in pubic rami fractures and in several surgical procedures [17,19– 21,37]. The coexistence of vascular variations originating from either the external iliac, inferior epigastric, or femoral vessels have been reported to require great attention from surgeons during surgical repair of inguinal and femoral hernias [11,21,32]. When present, the corona mortis can be damaged during fixation of the mesh on Cooper’s ligament in a laparoscopic herniorrhaphy, in which a staple or a tack is driven blindly into Cooper’s ligament [32]. This can cause uncontrollable bleeding, retroperitoneal hematoma, conversion to open surgery and reoperation [11]. During conventional hernia repair, damage to the anastomosis may result in significant hemorrhage, but this can usually be controlled by extension of the exposure [38]. An exposed corona mortis during oncological pelvic dissections also emphasizes the importance of exercising extreme caution when dissecting malignancies close to this collateral circulation [20]. However, even though corona mortis vessels were identified behind the superior pubic ramus in anatomical dissections during lymphadenectomies [20], another study reported that these corona mortis vessels did not seem to be as great a threat in clinical practice during these procedures [1].
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
G Model JINJ 7527 No. of Pages 7
6
B. Sanna et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx
Other serious complications reported include injury to the corona mortis following a TVT-Secur procedure with a U-approach, in which a tape is pushed behind the arc of the pubic bone, and the contact between the sharp, scalpel-like incisor and the vessel could cause damage to the corona mortis [39] resulting in severe bleeding into the space of Retzius [17]. Moreover, as the corona mortis is located behind the superior pubic ramus, it is susceptible to injury in scalpel dissections [37]. In addition, extreme care should be taken in the case of injury to the tendinous arch of pelvic fascia during paravaginal repair due to its close relation with the corona mortis [19]. To avoid iatrogenic errors in the abovementioned surgical procedures, radiologists should alert surgeons regarding the digital anatomical data of the corona mortis and illustrate the appropriate operation approach, reducing the risk of lacerating corona mortis vessels [37,40]. Our main findings revealed that the overall prevalence of the corona mortis in the population is high (49.3%), that it is more common in Asia (59.3%) than in Europe (42.8%) and North America (44.3%), and that a venous corona mortis (41.7%) is more commonly encountered than an arterial corona mortis (17.0%). In terms of geographical subgroups, the statistically significant difference between Asia, and Europe and North America, can be explained due to a more similar racial composition of the population and genetic similarities between Europeans and North Americans, as compared to Asians. Another explanation for the discrepancy in these findings could be the confusion in the literature over the term corona mortis and what it should include. The definition of the corona mortis used in our paper is the widely accepted definition [1,3] and should be consequently used in all populations to avoid unnecessary discrepancies in studies investing the corona mortis. Minimally invasive procedures, such as the insertion of the TVTSecur, have been linked to injury of the corona mortis due to the vessel’s position close to the place of the margin (25–50 mm from the pubic symphysis) [34]. The degree of individual variation concerning the position of the corona mortis, and the fact that both an arterial and venous corona mortis may be present, and in 17.9% of cases even bilaterally, highlights that a direct contact with the vessel is more probable in cases in which corona mortis vessels are located close to pubic symphysis. Moreover, surgeons should bear in mind that although the damage to an arterial corona mortis is more dangerous in terms of hemorrhage and hemodynamic instability, the identification of an active venous source of bleeding is more difficult and angiography cannot be used for this purpose [34]. A heightened awareness of the possibility of contact with the corona mortis should help reduce hemorrhage rates during minimally invasive procedures. Our meta-analysis was limited by the high amount of heterogeneity between studies. This can be explained by regional differences in the development of the vascular system, difficulties in dissecting the collapsed vessels in cadavers, and the limitations caused by the size of the vessels [7]. Small vessels often may be excluded, however, vessels smaller than 2 mm in diameter are especially important to map as they are poorly visible and harder to protect than larger vessels, increasing the risk of hemorrhagic complications in patients with small corona mortis vessels. There are few intraoperative studies investigating the corona mortis available in the literature. To be able to properly compare cadaveric and intraoperative studies, more intraoperative studies are needed. However, the large amount of cadaveric studies and their large sample sizes is a strength to our meta-analysis, as these studies are the gold standard for this type of anatomical investigation. In addition, future studies should be conducted in continents poorly represented by our meta-analysis, such as North America, South America, Africa, and Oceania to give a better
estimate of the true PPE of the corona mortis in various populations. Considering its clinical significance in numerous surgical procedures, the corona mortis, or Crown of Death, is an important anatomic variation that should be treated with the same respect as its name implies. Conclusions Our main findings revealed that the corona mortis was found to be present in nearly half of the general population, that a venous corona mortis is 2–3 x more prevalent than an arterial corona mortis, and that the corona mortis is more common in Asia than in Europe and North America. Based on our anatomical findings, we recommend an effort should be taken to identify and preserve the anastomotic vessels to prevent serious iatrogenic errors in surgical procedures in the retropubic region.
Acknowledgements Krzysztof A. Tomaszewski was supported by the Foundation for Polish Science (FNP) and the Polish Ministry of Science and Higher Education Grant for Young Scientists. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.injury.2017.12.007. References [1] Darmanis S, Lewis A, Mansoor A, Bircher M. Corona mortis: an anatomical study with clinical implications in approaches to the pelvis and acetabulum. Clin Anat 2007;20:433–9. [2] Drewes PG, Marinis SI, Schaffer JI, Boreham MK, Corton MM. Vascular anatomy over the superior pubic rami in female cadavers. Am J Obstet Gynecol 2005;193:2165–8. lu M, Uz A, Ozmen MM, Bozkurt MC, Erkuran C, Taner S, et al. Corona [3] Berberog mortis: an anatomic study in seven cadavers and an endoscopic study in 28 patients. Surg Endosc 2001;15:72–5. [4] Gilroy AM, Hermey DC, DiBenedetto LM, Marks Jr SC, Page DW, Lei QF. Variability of the obturator vessels. Clin Anat 1997;10:328–32. [5] Jakubowicz M, Czarniawska-Grzesiñska M. Variability in origin and topography of the inferior epigastric and obturator arteries. Folia Morphol (Warsz) 1996;55:121–6. [6] Namking M, Woraputtaporn W, Buranarugsa M, Kerdkoonchorn M. Corona mortis in thai. FASEB Journal 2006;20:A445. [7] Okcu G, Erkan S, Yercan HS, Ozic U. The incidence and location of corona mortis: a study on 75 cadavers. Acta Orthop Scand 2004;75:53–5. [8] Sarikcioglu L, Sindel M, Akyildiz F, Gur S. Anastomotic vessels in the retropubic region: corona mortis. Folia Morphol (Warsz) 2003;62:179–82. [9] Teague DC, Graney DO, Routt ML. Retropubic vascular hazards of the ilioinguinal exposure: a cadaveric and clinical study. J Orthop Trauma 1996;10:156–9. [10] Tornetta P, Hochwald N, Levine R. Corona mortis. Incidence and location. Clin Orthop Relat Res 1996;329:97–101. [11] Ates M, Kinaci E, Kose E, Soyer V, Sarici B, Cuglan S, et al. Corona mortis: in vivo anatomical knowledge and the risk of injury in totally extraperitoneal inguinal hernia repair. Hernia 2016;20:659–65. [12] Daeubler B, Anderson SE, Leunig M, Triller J. Hemorrhage secondary to pelvic fracture: coil embolization of an aberrant obturator artery. J Endovasc Ther 2003;10:676–80. [13] Garrido-Gómez J, Pena-Rodríguez C, Martín-Noguerol T, Hernández-Cortes P. Corona mortis artery avulsion due to a stable pubic ramus fracture. Orthopedics 2012;35:e80–82. [14] Henning P, Brenner B, Brunner K, Zimmermann H. Hemodynamic instability following an avulsion of the corona mortis artery secondary to a benign pubic ramus fracture. J Trauma Inj Infect Crit Care 2007;62:E14–7. [15] de Kleuver M, Kooijman MA, Kauer JM, Veth RP. Pelvic osteotomies: anatomic pitfalls at the pubic bone. A cadaver study. Arch Orthop Trauma Surg 1998;117:270–2. [16] Damsin JP, Lazennec JY, Gonzales M, Guerin-Surville H, Hannoun L. Arterial supply of the acetabulum in the fetus: application to periacetabular surgery in childhood. Surg Radiol Anat 1992;14:215–21.
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007
G Model JINJ 7527 No. of Pages 7
B. Sanna et al. / Injury, Int. J. Care Injured xxx (2017) xxx–xxx [17] Larsson P-G, Teleman P, Persson J. A serious bleeding complication with injury of the corona mortis with the TVT-Secur procedure. Int Urogynecol J 2010;21:1175–7. [18] Gobrecht U, Kuhn A, Fellman B. Injury of the corona mortis during vaginal tape insertion (TVT-SecurTM using the U-Approach). Int Urogynecol J 2011;22:443– 5. [19] Ersoy M, Sagsoz N, Bozkurt MC, Apaydin N, Elhan A, Tekdemir I. Important anatomical structures used in paravaginal defect repair: cadaveric study. Eur J Obstet Gynecol Reprod Biol 2004;112:206–13. [20] Pellegrino A, Damiani GR, Marco S, Ciro S, Cofelice V, Rosati F. Corona mortis exposition during laparoscopic procedure for gynecological malignancies. Updates Surg 2014;66:65–8. [21] Ramser M, Messmer AS, Zbinden I, et al. Incarcerated obturator hernialaparoscopic repair with intraoperative view of the corona mortis. J Surg Case Reports 2014;2014(8):rju081. [22] Moher D, Liberati A, Tetzlaff J, Altman D. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009;6(7):1000097. [23] Higgins JP, Green S. Cochrane Handbook for Systematic Reviews of Interventions Cochrane Book Series. THE COCHRANE COLLABORATION 1; 2011. [24] Henry BM, Tomaszewski KA, Walocha JA. Methods of Evidence-Based Anatomy: a guide to conducting systematic reviews and meta-analysis of anatomical studies. Ann Anat 2016;205:16–21. [25] Bible JE, Choxi AA, Kadakia RJ, Evans JM, Mir HR. Quantification of bony pelvic exposure through the modified Stoppa approach. J Orthop Trauma 2014;28:320–3. [26] Hong H, Pan Z, Chen X, Huang Z. An anatomical study of corona mortis and its clinical significance. Chin J Traumatol: Zhonghua chuang shang za zhi 2004;7:165–9. [27] Kacra BK, Arazi M, Cicekcibasi AE, Buyukmumcu M, Demirci S. Modified medial stoppa approach for acetabular fractures: an anatomic study. J Trauma Inj Infect Crit Care 2011;71:1340–4.
7
[28] Karakurt L, Karaca I, Yilmaz E, Burma O, Serin E. Corona mortis: incidence and location. Arch Orthop Trauma Surg 2002;122:163–4. [29] Kawai K, Honma S, Koizumi M, Kodama K. Inferior epigastric artery arising from the obturator artery as a terminal branch of the internal iliac artery and consideration of its rare occurrence. Ann Anat 2008;190:541–8. [30] Mahato NK. Retro-pubic vascular anomalies: a study of abnormal obturator vessels. Eur J Anat 2009;13:121–6. [31] Pungpapong S, Thum-umnauysuk S. Incidence of corona mortis; preperitoneal anatomy for laparoscopic hernia repair. J Med Assoc Thai 2005;88(Suppl. 4): S51–3. [32] Rusu MC, Cergan R, Motoc AG, Folescu R, Pop E. Anatomical considerations on the corona mortis. Surg Radiol Anat 2010;32:17–24. [33] Smith JC, Gregorius JC, Breazeale BH, Watkins GE. The corona mortis, a frequent vascular variant susceptible to blunt pelvic trauma: identification at routine multidetector CT. J Vasc Interv Radiol 2009;20:455–60. [34] Stavropoulou-Deli A, Anagnostopoulou S. Corona mortis: anatomical data and clinical considerations. Aust New Zeal J Obstet Gynaecol 2013;53:283–6. [35] Baena G, Rojas S, Peña E. Corona mortis: anatomical and clinical relevance and occurrence in a sample of the colombian population. Int J Morphol 2015;33:130–6. lu M, Uz A, Ozmen MM, Bozkurt MC, Erkuran C, Taner S, et al. Corona [36] Berberog mortis: an anatomic study in seven cadavers and an endoscopic study in 28 patients. Surg Endosc 2001;15:72–5. [37] Al Talalwah W. A new concept and classification of corona mortis and its clinical significance. Chin J Traumatol: Zhonghua chuang shang za zhi 2016;19:251–4. [38] Gaines RD. Complications of groin hernia repair: their prevention and management. J Natl Med Assoc 1978;70:195–8. [39] Castellani D, Saldutto P, Galica V, Masciovecchio S, Galatioto GP, Vicentini C. Transobturator surgery in presence of corona mortis: a study in 13 women. Urol 2016;83:200–3. [40] Han Y, Liu P, Chen C, Duan H, Chen L, Xu Y, et al. A digital anatomical study of the corona mortis in females. Minim Invasive Ther Allied Technol 2016;26:111–8.
Please cite this article in press as: B. Sanna, et al., The prevalence and morphology of the corona mortis (Crown of death): A meta-analysis with implications in abdominal wall and pelvic surgery, Injury (2017), https://doi.org/10.1016/j.injury.2017.12.007