Traumatic Internal Mammary Injury— A Literature Review and Guide for Management


 BackgroundTraumatic internal mammary artery (IMA) injury is a rare but clinically significant injury since it is linked to various troublesome conditions. In this review, we aim to provide a complete description and a practical strategy for managing this injury.MethodsPubMed, Embase, and the Web of Science were searched from 1960 to 2019. We also reviewed IMA injuries in our hospital. Only traumatic IMA injuries were included. The demographics, diagnostic procedures, treatments and outcomes were recorded and analyzed.ResultsThere were a total of 72 IMA injury cases, including 4 in our hospital. The injuries were predominantly in males (79.2%), were on the left side (66.7%), were considered blunt injuries (63.9%), and featured mediastinal hematoma (62.5%). Computed tomography (CT) has been used to diagnose IMA injuries in recent years. Most patients could be managed by nonsurgical treatments (61.2%). The outcome for IMA injury is very good, with low morbidity and mortality rates.ConclusionsIdentifying traumatic IMA injuries requires alertness from the physicians. An appropriate management chosen according to the patient’s individualized condition could lead to a satisfying result.Trial registration: not applicable


Abstract Background
Traumatic internal mammary artery (IMA) injury is a rare but clinically signi cant injury since it is linked to various troublesome conditions. In this review, we aim to provide a complete description and a practical strategy for managing this injury.

Methods
PubMed, Embase, and the Web of Science were searched from 1960 to 2019. We also reviewed IMA injuries in our hospital. Only traumatic IMA injuries were included. The demographics, diagnostic procedures, treatments and outcomes were recorded and analyzed.

Results
There were a total of 72 IMA injury cases, including 4 in our hospital. The injuries were predominantly in males (79.2%), were on the left side (66.7%), were considered blunt injuries (63.9%), and featured mediastinal hematoma (62.5%). Computed tomography (CT) has been used to diagnose IMA injuries in recent years. Most patients could be managed by nonsurgical treatments (61.2%). The outcome for IMA injury is very good, with low morbidity and mortality rates.

Conclusions
Identifying traumatic IMA injuries requires alertness from the physicians. An appropriate management chosen according to the patient's individualized condition could lead to a satisfying result.
Trial registration: not applicable Background Internal mammary artery (IMA) injury is uncommon among chest traumas. However, this injury is known as a red ag to trauma physicians because of its potential for tremendous life-threatening bleeding (1). In addition, sometimes these injuries could be insidious at rst and therefore masked by other more severe injuries (2).
The rst case report can be traced back to 1963 (3); however, these injuries have not been well discussed in the literature since then. The aim of this study is to provide a comprehensive review of IMA injuries, review our experiences, and establish a clinical guide.

Methods
PubMed, Embase, and the Web of Science were searched from 1960 to 2019 for available articles using the following keywords: ''internal mammary artery injury,'' ''internal thoracic artery injury'', "thoracic injuries", "mammary arteries" and "chest trauma". We focused on trauma cases in this review, so iatrogenic IMA injuries were excluded. Articles were also excluded if general data, the clinical course and the diagnostic tools could not be retrieved from the full text or the abstract.
Two trauma surgeons independently reviewed all studies. Discrepancies between the two reviewers were resolved by a third author.
We also reviewed the trauma registry database of our hospital, which is a level I trauma center, and collected all IMA injuries from 2009 to 2018. For all the included studies and patients, the following variables were recorded: patient age, sex, injury mechanism, lesion side, clinical presentation, diagnostic methods, treatment, and outcome.

Results
During the 50-year period, 38 articles were found, including 4 case series and 34 case reports . The largest case series had 18 patients. Among the included articles and our database, there were a total of 72 IMA injuries, four of which were from our hospital. The mean patient age was 40.9 years (range 16-86 years), and there was a male predominance (79.2%).
The general data are listed in Table 1. Forty-six patients (63.9%) had blunt injuries, of which tra c accidents were the most common etiology. Twenty-six patients (36.1%) had penetrating injuries, of which stab injury was the most common cause. The left side (66.7%) was involved much more often than the right (16.7%) and bilateral (16.7%) sides. At least 58% of patients had associated injuries. Adjacent structures, such as the heart, lung, nearby vessels, sternum, and rib, were more likely to be involved, while some patients had multisystem injuries. Table 2, the most common recorded presentation was mediastinal hematoma (45 patients, 62.5%), followed by hemothorax (33 patients, 45.8%). Among the patients with mediastinal hematoma, nearly one quarter (11/45, 24.4%) had extracardiac tamponade. Fifty-three percent of the cases were reported taking plain chest lms. Mediastinal widening was a common nding on plain chest lm (63.2%) but was not always present. Based on the retrieved studies that clearly mentioned diagnostic modalities, most of the patients received computed tomography (75.5%), angiography (73.6%), or both (26/53, 49.1%) to identify the lesion sources. The types of IMA injuries according to angiography or surgical ndings were described in 57 patients. Transection (38.6%) was the most common injury pattern, followed by contrast extravasation (35.1%) and aneurysm/pseudoaneurysm (17.5%). Arteriovenous stulas (AVFs) with various adjacent veins were not common but were reported in 5 cases (8.8%) (3,6,19,20,30).

As shown in
The management strategies and outcomes are shown in Table 3. Nonsurgical treatment was applied in 61.2% of the patients, including 55.6% who underwent transarterial embolization (TAE) and 5.6% who received conservative treatment. The rest of the patients required surgical intervention. Sternotomy was the most frequent approach, while few patients were managed by antero-lateral thoracotomy incision.
To clarify the decision-making process when confronting IMA injury, fty-three patients with a detailed clinical course were selected and are summarized in Figure 1 (1-17, 19-35, 37, 38). Five patients were sent to the operating room directly, while 14 patients underwent surgery after the CT results were reviewed.
Among the 34 patients receiving angiography, only 27 patients were managed with embolization, and the other seven patients still underwent surgery.

Discussion
IMA originates from the subclavian artery and comprises the anterior intercostal, perforating, musculophrenic, superior epigastric, and pericardiophrenic branches. Injury to the IMA and/or its branches is rare in chest trauma, but the anatomical proximity to the heart could result in signi cant and life-threatening bleeding. In addition, bleeding con ned to the mediastinum may result in extracardiac tamponade or the accumulation of blood in the thoracic cavity, resulting in tension hemothorax; both of these conditions could further compromise hemodynamic stability. Early detection is challenging but is very important in clinical practice.
Most chest trauma patients receive a plain chest lm rst. However, for IMA injury, an unacceptably high proportion of patients would be missed since there are no signi cant ndings. Before multidetector CT was prevalent, angiography was arranged for patients with a high suspicion of mediastinal vessel injury. In our review, eight patients underwent angiography without CT, all of whom were from studies before 2000 (3,6,15,17,19,30,34,37). After 2000, although diagnosis and treatment could be performed simultaneously by angiography, most physicians chose contrast-enhanced chest CT as the diagnostic tool because of its easy accessibility and ability to provide more comprehensive information. For blunt chest trauma, the National Emergency X-Radiography Utilization Study chest algorithm (NEXUS-chest) has already provided well-validated and practical recommendations for the implementation of chest imaging (39). For penetrating chest trauma, if an IMA injury is suspected according to the depth and route of injury, a contrast-enhanced chest CT scan is the reasonable rst choice as long as the patient's condition is suitable for the exam.
The treatments for IMA injury include observation, TAE, and surgery. Four patients in our review recovered after observation, and all were from the same series (36). In the series from Whigham et al., angiography was performed in all of their 18 patients, but four of them did not receive embolization since no active bleeding was identi ed. According to our review, TAE for IMA bleeding was rst described in 1982 (15). Since the technique and image quality have made large progress in recent decades, TAE has become the mainstay choice in the literature for managing IMA bleeding due to its less invasive nature and equivalent success rate for hemostasis compared to surgery.
Shock remained the most important reason for surgery. Five patients in our review were operated on directly without CT or angiography (2,4,14,17). Four patients were in shock status before the operation, and the other patient was sent back to the operating room due to massive hemothorax after receiving laparotomy for intra-abdominal bleeding. Fourteen patients underwent surgery after the initial CT scan, including 4 patients for shock (1,12,28), three for extracardiac tamponade (10,16,23), two for massive hemothorax(1), two for both extracardiac tamponade and hemothorax (1,24), one for a large aneurysm (25), one for a foreign body (5), and one for persistent bleeding after pericardiotomy (11). Seven patients underwent surgery after angiography. Three were for AVF (3,6,19), two for shock (17,22), one for combined subclavian and vertebral artery injury (27), and one for diagnosis only (37).
The outcome of IMA injury is good. Although the overall mortality rate seems high in our review, only one death was caused by uncontrolled exsanguination due to IMA, while the others were caused by severe associated injuries or organ failure. Most patients had an uneventful course after hemostasis was achieved. In fact, no patients with isolated IMA injury had complications. All reported complications were related to the respiratory system and occurred in those with associated injuries.
Notably, not all IMA injuries were identi ed at the beginning. Nearly 10% (7 patients) had a delayed presentation (6,8,19,20,25,30,37), especially those with AVF (4 in 5 patients). The duration from the trauma to the delayed diagnosis ranged from 1 week to 10 months. Little is known about these injuries were not identi ed at rst presentation, but we believe that sometimes subtle vessel injuries such as AV shunting do need time to become symptomatic and evident. Figure 2 is an integrative algorithm for IMA injury that we created after reviewing all of the literature. After performing the initial assessment and resuscitation according to the principles of Advanced Trauma Life Support 10 ed., extremely unstable patients should undergo surgical treatment promptly. For those who could be stabilized, the NEXUS CT-major or NEXUS CT-all criteria should be applied to decide the next step for blunt injuries, and CT is recommended for penetrating injuries if an IMA injury is suspected. If there is no evidence of active bleeding, conservative treatment could be considered, but these patients need close observation. If there is evidence of active bleeding, i.e., contrast extravasation or aneurysm/pseudoaneurysm, TAE could be considered rst if the patient is clinically stable. If the following indications exist, surgical exploration is preferred prior to nonsurgical treatment: massive hemothorax, extracardiac tamponade, foreign bodies that need to be removed, di cult approach for TAE (e.g., AVF or large aneurysm), or concomitant injuries that need surgery immediately (e.g., cardiac rupture).

Conclusions
IMA injury is uncommon but should be kept in mind from the initial to delayed phase when managing chest trauma to avoid unfavorable consequences. CT is thus far the main diagnostic tool for IMA injury. TAE has a high success rate in stopping bleeding in stable patients, while surgery still plays an important role in unstable patients or in certain speci c circumstances that could not be resolved by nonsurgical treatments. Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests Funding This research did not receive any speci c grant from funding agencies in the public, commercial, or notfor-pro t sectors. Summary of 53 patients whose treatment course was described in the literature. For these 53 patients, 27 were managed by transarterial embolization (TAE), and 26 underwent surgery. Among the 26 patients who underwent surgery, 5 underwent surgery directly, 14 underwent surgery after CT, 5 underwent angiography, and the remaining 2 underwent surgery following both CT and angiography. Proposed algorithm for the management of IMA injury. Immediate operation is indicated for patients with extremely unstable hemodynamics. The NEXUS CT-Major or All criteria were applied as selection criteria for CT in patients with blunt chest trauma (39). TAE should be considered for stable patients with active bleeding, while surgery should be considered for several situations, as shown in the algorithm. NEXUS CT-Major = 1) distracting injury; 2) chest wall tenderness; 3) sternal tenderness; 4) thoracic spine tenderness; and 5) scapular tenderness. NEXUS CT-All = CT Major criteria + rapid deceleration mechanism. Patients having 2 or more criteria present should obtain chest CT. Patients having 1 criterion present should consider CT. Patients having all criteria absent are safe to forego chest CT.