1.Introduction
Descending Necrotizing Mediastinitis (DNMs) was first described in 1938 by Pearse HE [1] as a serious infectious regional association at neck area spreading to chest, and death mostly due to untreated. In 1938 Estrera AS [2] gave the criteria for the diagnosis of ARB still applied to this day. Despite progress in treatment and resuscitation so far, mortality is as high as 60% in a number of reports. The patient died mainly from severe infections or acute bleeding due to damage to large vessels. Infectious control in DNMs need proper attention apart from emergency surgery for purulent drainage [1-3].

In Vietnam nowadays, there are not many studies on DNMs. Some studies by Nguyen Duc Chinh et al at Viet Duc hospital (2001 and 2017) [4], Nguyen Cong Minh (2014) at Cho Ray hospital, overall mortality ranges from 17% to 35% [5]. However, not so many in-depth studies on DNMs are particularly related to microbiological and antibiotic treatment, so the main purpose of this study was to learn the microorganisms isolated and the antibiotic treatment strategies.

2. Subjects and Methodology

2.1. Objects and Research Methods

Subjects

Patients diagnosed with DNMs due to perforation of esophagus are treated at Viet Duc hospital, including death, from 2016 to 2019, regardless of age.

Selection criteria:

Criteria diagnosis of DNMs according to Estrera (1983) [2]:

Classification according to Endo S (1999) [6]:

Sample size: all patients during the study period were enrolled,

Exclusion criteria:

The cases of DNMs are not caused by esophageal perforation.

Died without surgery or forensic examination

2.2. Research Methods

Retrospective combined with prospective study, clinical description, vertical follow-up, no comparison.

Design medical records to collect information

Cultured microorganisms are identified by the automatic method

Antibiotic therapy by experience and by antibiogram results.

3. Results

3.1. Some clinical and subclinical features

Relationship between ages and sex are shown in the table 3.1

Average ages are 56,3 ± 8,6.

Table 3.1: Characteristics related to age and sex

The cause of perforation by injuries: The majority of injuries account for 70% (28/40), the pathology is 30%, and there is no cause of medical intervention.

In the cause of foreign bodies, the main bones are 23/26 accounting for 88.5%, of which chicken bones are 12/23 (52.2%), fish bone are 9/23 (39.1%).

The cause of perforation by pathology: Boehaave mainly accounts for 5/12, unknown 4/12.

Classification according to Endo S

Management: one by preservative treatment; 39 by emergency surgical interventions

Table 3.2: Classification by Endo S

The procedures were drainage 82.1%, chest opening 17.9%. Drainage includes: cervical drainage 40.6%; Chest Drainage 31.2%, Neck Drainage + Chest 28.1%. Open chest 7/39 including open surgery (05) and Video Assisted Thoracothomy (02)

3.2. Microbiological test results:

24/40 cases of microbial isolation, accounting for 60%

                                              Table 3.3: Microbiological isolation results

The bacteria can only be isolated once including: Streptococcus contell, Streptococcus angirosus, Streptococcus mitis, Staphylococcus aureus ,Acinetobacter, Burkhoderia cepacia,Enterococcus,Citobacter diversus,Prevotella intermedia  anaerobic.

Fungi was isolated in 6/24 accounting for 25%.

Antibiogram: In addition to Staphylococcus Epidermidis, which is not drug resistant, the bacteria isolated are resistant, in which Acinetobacter baumanii, Citobacter, Burkholderia cepacia, and Peptostreptococcus are resistant to many.

3.3. Outcomes 

Antibiotic therapy:

                                                             Table 3.4:  Antibiotic used by experience

These antibiotics were used firstly by experience. However, the results of the antibioticgram were shown that they were not resistant and continued to be used.

Complications: 7/40  postoperative complications were noted, accounting for 17.5%

                                                 Table 3.5: Postoperative complications

The mortality was noted in 3/40, accounting for 7,5% including one serious bleeding from large vessal injury, one septic shock, one multiple failure. 

4. Discussion 

4.1.Some characteristics of objectives

Age: The mean age of patients with DNMs in our study was 56.3 ± 8.6, while in the study of other authors Ricardas J was 55.3 ± 15.4, Sofia Arizaga was 40 years old. Thus, the age of patients in our study is older [4.7].

Gender: In our study, men account for 4.7: 1 higher than women, consistent with other authors such as Ricardas J is 2.7: 1, Nguyen Duc Chinh 3: 1. [7-9]

The causes of perforation: Most of international authors have reported that  the iatrogenic esophageal perforation accounted for 70%. Amudhan A's study of 726 long-term patients from 1990 to 2003 [10] in India showed that esophageal perforation by medical intervention accounted for 19% (about 7% to 33%), trauma accounted only for about 7% (from 0% - 33%), of Gupta NM [11] accounts for 70% due to medical intervention. In our study, the leading cause of trauma accounted for 70%, especially the cause of bones injured accounted for 88.5%, no case related to medical intervention. In the study of Nguyen Duc Chinh [4], there were 02 related cases after the cervical spine vis-brace surgery. Thus in Vietnam, the rate of iatrogenic esophageal perforation is very low, but it can occur, it is necessary to pay attention to the spine surgery with implant.

Management: one preservative treatment; 39 were treated surgically on emergency Including drainage 82.1%, chest opening 17.9%. 

Drainage includes: cervical drainage 40.6%; Chest drain 31.2%, Neck + Chest drainage 28.1.

4.2 Microorganisms

According to F. ethui, Nguyen Duc Chinh [9,12] bacteria isolated are 100% poly-bacteria, combining both aerobic and anaerobic, no case of isolating only one  bacteria. The results of microbiological examination help physician to have appropriate antibiotic treatment strategies.

According to Luis Marcelo (2003) [13] the most common bacteria are throat hemolytic streptococci (beta-hemolytic oral Streptococcus). Other common bacteria are Prevotella, Peptostreptococcus, Fusobacterium, Veillonella, Actinomyces, Bacterioides, Staphylococcus, alpha-hemolytic Streptococcus.

In the study of Weaver E (2010) [14] microorganisms combine both anaerobic and aerobic and are the most common oral group, in which aerobic is hemolytic streptococcus alpha haemolytic Streptococcus, Staphylococcus aureus, Klebsiella pneumonia. The common anaerobic bacteria are Peptostreptococcus, Bacteroides fragilis, Prevotella Porphyromonas. In the study of Muhammad Arza (2016) [15], the main bacteria isolated were Pseudomonas aeruginosa, Enterobacter aerogenes, Acinetobacter baumannii. According to Ochi N (2018) [16] Gram (+) bacteria isolated were Streptococcus pyogenes, Streptococcus constellatus, Fusobacterium necrophorum, and Peptostreptococcus sp.

In our study, the combination of both aerobic and anaerobic microorganisms was 100%. (Table 3.3). In 24 positive cases: 10 patients had one bacteria, 9 patients combined 2 bacterias (including 8 patients with a combination of anaerobic bacteria Peptostreptococcus), 4 patients combined 3 bacterias, one patient has 4 types of pathogenic bacteria.

The study of microorganisms in our series is also consistent with the authors showing that pathogenic microorganisms are a combination of both aerobic and anaerobic (100%). Gram (+) aerobic bacteria accounted for the majority of 75.9%, the most was Streptococcus species 37.5%. Anaerobic bacteria Gram (-) accounted for 17.2%, the most was Klebsiella pneumonie accounting for 12, 5%, Pseudomonas aeruginosa 12,%. Gram (+) anaerobic bacteria: Peptostreptococcus accounts for 8.3%.

4.3. Antibiotic treatment

Septic shock is one of the high risk factors for death. Reported by Nguyen Cong Minh (2012) [5] all cases of DNMs with septic shock were fatal. The antibiotic treatment must be early and the antibiotic is broad spectrum, but the result of late culture should be used by experience. Nguyen Duc Chinh (2015) [4] reported the mortality 16.7%, due to serious bleeding and septic shock (multi-organ failure) despite taking early and strong antibiotics once at admission. According to S. Diamantis [17] when the results of microbiological studies were not available, the authors recommended using intravenous antibiotics starting with vancomycin and unasyn. Ochi N [16] and Japanese colleagues suggested that the first choice antibiotic was 600 mg clindamycin bid and 4. 5 g of tazobactam / piperacillin hydrate intravenously and most patients were well controlled the infection.

Our study results showed that in addition to Staphylococcus Epidermidis, which is not drug resistant, the isolated bacteria are resistant to drugs, in which Acinetobacter baumanii, Citobacter, Burkholderia cepacian, Peptostreptococcus were resistant to many antibiotics. The experiences with the ultimate combination of strong and broad-spectrum antibiotics, Cephalosporin (3rd generation) and Carbapenem, and Metronidazole have been shown to be all effective and we found that it was not necessary to change antibiotics compared with antibiotic results later (Table 3.4).

5. Conclusion

Microbiological characteristics in the study showed that the combination of anaerobic and aerobic isolated were 100%, in which the aerobic Gram (+) was the majority. The experiences with the ultimate combination of strong and broad-spectrum antibiotics, Cephalosporin (3rd generation) and Carbapenem, and Metronidazole have been shown to be all effective.

Conflict of interest: none