Case Report
Acute Descending Necrotizing Mediastinitis: Spectrum of Management and Role of Video Assisted Thoracoscopic Surgery
Habib Khan1, Sanjay Asopa1, Salil Nair2, Eunan Tiernan3 and Khalid A Amer1*
1The Cardiovascular & Thoracic Centre, University Hospital Southampton, UK
2Department of Otolaryngology, University Hospital Southampton, UK
3Department of Plastic Surgery, Salisbury District Hospital, UK
*Corresponding author: Khalid A Amer, The Cardiovascular & Thoracic Centre, University Hospital Southampton NHS Foundation Trust, Southampton General Hospital, Tremona Road, Southampton SO16 6YD, UK
Published: 25 May, 2017
Cite this article as: Khan H, Asopa S, Nair S, Tiernan
E, Amer KA. Acute Descending
Necrotizing Mediastinitis: Spectrum
of Management and Role of Video
Assisted Thoracoscopic Surgery. Clin
Surg. 2017; 2: 1482.
Abstract
A thoracic surgeon in the UK is likely to come across Descending Necrotizing Mediastinitis once
a year in a busy tertiary referral hospital. This is a rare but life-threatening complication of oropharyngeal
sepsis. It carries a poor prognosis if recognised late. In this report we selectively describe
five cases treated differently, with emphasis on Computed Tomography guidance to bespoke
surgical drainage of pus. We specifically discuss two aspects of our experience with these patients.
Firstly the spectrum of surgical management and role of Video Assisted Thoracoscopic Surgery
(VATS). Secondly we discuss “dysphagia”, which is invariably a presenting symptom/ complication,
its progression, and natural course. We also emphasise the important role of different teams on the
Intensive Therapy Unit as a major cause of these patients surviving the gruesome infection.
Keywords: Video assisted thoracoscopic surgery; Necrotizing Mediastinitis; Spectrum
Introduction
Descending necrotizing mediastinitis (DNM) is a rare complication of oropharyngeal and
odontogenic infections (tonsilitis, pharyngitis and tooth abscess). It usually affects young adults,
who were previously fit and healthy. Specific aetiologies have not been recognised, although patients
with insulin-dependent diabetes and alcoholics are more commonly affected [1]. Mixed aerobic
and anaerobic organisms are usually the culprit. A special form of DNM caused by fusobacterium
necrophorumis known as Lemierre’s syndrome, which presents as acute oropharyngeal infection
with life threatening septic thrombophlebitis of the internal jugular, and or subclavian veins and
metastatic abscesses [2].
The point of origin of the infection is almost always located deep in the cervical area. Among
these, periodontal disease (tooth abscess) constitutes 41%, while pharyngeal abscess is the second
most frequent aetiology [3]. The infection spreads along the cervical spaces into the retropharyngeal
space to reach the posterior mediastinum. The submental and submaxillary abscesses, however,
spread posteriorly towards the vascular space, and thereby may extend into the anterior mediastinum
[4]. Grodinsky and Holyoke presented the concept of the three fasciae and fascial spaces and showed
how an infection in these areas is likely to gravitate to the mediastinum [5].
Odynophagia (painful swallowing) and later dysphagia (resulting in regurgitation and
aspiration) is invariably reported in most published series. Dysphagia continues into the recovery
period posing a challenge to nutrition in the severely debilitated patient. It also predisposes to
aspiration, lung consolidation and lung abscess formation. The full pathology of dysphagia in DNM
is not fully understood.
In this report we selected five cases encountered in the last 10 years to describe the Southampton
experience in management of DNM. The emphasis is on the varied clinical picture, progression and
the bespoke surgery designed. We also discuss the invariable symptom of dysphagia in this group
of patients.
Case Presentation
Case 1
A 17-year-old female was admitted with a four days history of sore throat, difficulty swallowing and fever. Haematological investigations revealed raised white cell count (WCC=16 × 109/L), C-reactive protein (CRP=300 mg/L) and
positive Ebstein Barr virus serology. She was treated for tonsillitis
with intravenous Ceftriaxone 1 gm 8 hourly (tds), but failed to
respond. Subsequent CT scan of the neck and thorax demonstrated
a pharyngeal perforation with extensive retropharyngeal abscess
extending from the suprahyoid region to the aortic arch, air in the soft
tissues of the neck and prevertebral fascia, and fluid tracking down
into superior mediastinum around the great vessels. Her condition
deteriorated necessitating ventilatory and inotropic support in the
Intensive Therapy Unit (ITU).
In view of this, the patient underwent surgical exploration by
the Ear Nose and Throat (ENT) surgeons of the left anterior cervical
triangle as well as a right anterior mediastinotomy (Chamberlain
approach) by thoracic surgeons. Through the neck incision
oedematous tissue, pus and necrotic debris were encountered in the
plane lateral to the thyroid. Necrotic tissue was debrided, the loculi
were broken and pockets of pus in the retro-pharyngeal and retroesophageal
spaces were drained. Surgical toilet with Aqueous Iodine
and 3% hydrogen peroxide (H2O2) was performed and a 28 F Argyll
tube placed behind the oesophagus from a cervical approach. Through
the anterior mediastinotomy approach the pleura was entered
medially and dissected off the superior vena cava (SVC), where more
pus and necrotic tissue were encountered. This was debrided and a
16 F Redivac catheter drained the space. A right intercostal tube was
also placed through a different stab wound to drain a pleural effusion.
Microbiological analysis revealed mixed culture and the patient was
started on intravenous Clindamycin 400 mgs tds, Imipenem 2 gms
tds, and Metranidazole 400 mg tds.
Despite initial improvement in the post-operative course, the
patient relapsed with septicaemia on the 4th post operative day
requiring inotropic support. A repeat CT scan demonstrated residual
collection in the right perivascular space anterior to the SVC, left
parapharyngeal space and anterior to the aortic arch and right
internal jugular vein thrombosis (a feature of Lemiere’s disease). The
neck was subsequently re-explored via the initial cervical incision.
With blunt dissection, a track was created behind the pharynx to the
right and a corrugated Yates drain was placed behind the pharynx,
across the midline. Next the left cervical incision was joined to the
right anterior mediastinotomy incision by a J-shaped manubriotomy
(leaving the sterno-manubrial joints intact), giving excellent exposure
to the retro-manubrial space. The pockets of pus in the space around
the SVC and the aorta were drained, necrotic tissues debrided and
corrugated Yates drain left in situ. The surgical wounds were left open.
The patient’s post-operative course was complicated by persistent
swinging pyrexia and increasing inotropic requirement. Repeat CT
scan showed a collection in the right retropharyngeal space and at
the base of the skull, requiring drainage by right cervical incision and
peroral incision. A left pleural effusion required tube thoracostomy
drainage. She underwent a percutaneous endoscopic jejunostomy
(PEJ) for feeding. Tracheostomy was required for the protracted
weaning off the mechanical ventilation. Following this, the patient
made slow but progressive improvement and was eventually
weaned off the ventilator. Three months later her sternal wound was
closed using a left pectoralis major flap from which she recovered
uneventfully and was discharged home. Dysphagia was encountered
immediately after weaning off the ventilator, and the Speech and
Language Team (SALT) were involved in her care at an early stage.
The patients’ dysphagia recovered over a period of six months and
recent video assisted fluoroscopy revealed a residual left vocal cord
palsy with intact swallowing reflex. She was well at 2 years follow up.
Case 2
A 65 yr old lady was referred with a two week’s history of sore
throat, acute epiglottitis, anterior neck swelling and odynophagia
progressing to dysphagia. Early in the course of her illness she started
discharging pus through her mouth. She was treated with intravenous
antibiotic (Tazocin 4.8 grams tds.) to which she had a good
response. Her CT scan demonstrated an extensive prevertebral right
retropharyngeal abscess throughout the superior mediastinum (Figure
1C). A barium swallow excluded oesophageal perforation. Follow up
CT scan a week later demonstrated a considerable improvement in
the collections in the neck and superior mediastinum with reduced
peri-oesophageal oedema. Microbiological analysis on blood, sputum
and pus culture were negative. In view of being symptom free and
the spontaneous drainage of the posterior pharyngeal abscess she was
treated conservatively with intravenous and later oral antibiotics. Her
dysphagia improved over a period of two weeks and was tolerating
fluid and solids of all consistencies without regurgitation. After two
weeks in hospital she was discharged home symptom free, without
surgical intervention.
Case 3
A 67 yrs old lady was admitted to the hospital with a 2-days’
history of sore throat, odynophagia and subsequent right sided
anterior neck swelling. Consequently the odynophagia progressed
to complete dysphagia with nasal regurgitation. On examination
she was pyrexial at 38.2°C, with an evident 3 cm x 4cm swelling in
the right anterior cervical triangle. A fine nasal endoscopy (FNE)
revealed a right sided parapharyngeal swelling obliterating the
aryepiglotic fold. A CT scan illustrated marked abscess in the right parapharyngeal space, extending into the right submandibular gland.
Intravenous antibiotic therapy with Ceftriaxone 2 grams once daily
(od) and Metronidazole 500 mgs tds failed to control the sepsis. A
repeat CT scan demonstrated the right parapharyngeal collection
extending from the base of the skull to the superior mediastinum,
into the posterior mediastinum to the level of the carina. There were
also bilateral pleural effusions.
In view of this, the patient underwent exploration of her right
cervical triangle. One high and one low stab incisions in the midline
were fashioned. The abscess cavity was explored superiorly upto the
base of the skull, posteriorly to the posterior triangle, anteriorly to
the midline in front of the trachea and laterally to the ramus of the
mandible. The anterior mediastinum was approached through the
low stab incision via the pretracheal plane. This was communicated
with the right lateral and posterior mediastinum via a separate plane
lateral to the carotid sheath. A 32F Argyll drain was introduced from
the neck wound in the left paratracheal space all the way into the
mediastinum blindly. A 28F Argyll intercostal drain was inserted
into the right chest. A tracheostomy was fashioned the post-operative
recovery was slow and protracted. Naso-gastric tube was used to feed
the patient as his swallowing was impaired. Speech video fluoroscopy
demonstrated weakness in the right pharyngeal wall to be the cause of
dysphagia. The tracheal tube was downsized and removed when the
patient swallowing improved. She was discharged home after 4 weeks
in hospital and was well at 6 months follow up. Her swallowing had
returned to normal.
Case 4
A 42 year old gentleman very fit previously presented to the
ENT department with severe tonsillitis, and odynophagia. His
inflammatory markers were significantly high, WCC=17 × 109/L,
CRP=728 mg/L. His HIV status was negative. Initial intravenous
antibiotics included; Tazocin 4.5 grms tds, Clindamycin 12 gm
qds, Clarithramycin and metronidazole 1 gm bd. His sepsis was
not controlled, therefore; under local anaesthesia a retropharyngeal
abscess was drained via bilateral neck incisions. There was no Quincy.
3 corrugated drains were left on the right side and 2 similar drains
on the left side of the neck. Group G Streptococci were isolated from
drained pus. The wounds and drains were reviewed in theatre 5 days
later as he was not making progress. A CT was ordered and this lead
to the diagnosis of DNM (Figure 1A). The mediastinum showed
collection of pus, oedema and gas formation, together with bilateral
effusions and lung patchy consolidation. An echocardiogram revealed
a moderate pericardial effusion that did not require drainage. A left
chest drain was inserted but without much improvement. Two days
later he underwent right thoracotomy and opening of mediastinum
from neck to diaphragm. Originally the operation was planned as
VATS, but sepsis, hypoxia and retention of CO2 necessitated high
inflation pressures making lung isolation impossible, therefore a
single lumen Endotracheal tube was used. After initial improvement
in the Intensive care unit and substantial reduction in WCC and CRP
he relapsed one week later, with a temperature of 37.9°C and rising
markers. The ENT team reviewed the neck drains, and a second CT
was ordered. The latter showed a new collection of pus anterior to
the arch of the aorta and to the back of the left hilum, not drained by
the left chest drain or the new right mediastinal drains (Figure 1B).
Further left sided surgery was necessary. VATS would have been the
preferred approach; however for the same reasons as in the previous
operation isolation was impossible. Initially a limited left anterior
mediastinotomy (Chamberlain approach) was tried to gain access to
the arch of aorta. The high inflation pressures made it impossible to
retract the apex of the lung down to expose the aorta. The wound was
closed and a proper posterolateral left thoracotomy was opened. The
pleural reflection over the aorta between phrenic and vagus nerves was
opened from below over the main pulmonary artery, up over the arch
and all the way over the neck vessels. The crossing superior intercostal
vein was thrombosed. The incision was continued down behind the
hilum to the level of diaphragm, disconnecting all bronchial vagal
branches. The finger was used to open all loculations of pus. A 36F
posterior and 32 F anterior drains were inserted in the left chest.
His subsequent recovery was slow, but finally was extubated
without the need for a temporary tracheostomy. Because we had
some experience with dysphagia by this time, the SALT team were
involved early, and persevered with Naso-Gastric Tube feeding,
rather than fashioning a PEJ. He was discharged home after 4 weeks
in hospital and it took over two months for his swallowing to come
back to normal.
Case 5
An athletic 48 year old gentleman presented with sore throat,
odynophagia and subsequent right sided anterior neck swelling. CT
scan of his head, neck and chest revealed extensive neck and mediastinal
abscess and bilateral empyema and a diagnosis of DNM (Lemierre’s
Disease) was made. Neck abscess was drained and two chest drains
placed to drain 1.5 litres of pus. Blood Cultures grew Fusobacteria
and antibiotics were changed to Tazocin and Metronidazole. Duplex
studies confirmed thrombosis of his left subclavian vein and he was
anticoagulated with short acting Heparin and Warfarin. He was then
transferred to the ITU for septic deterioration in preparation for
emergency surgical drainage. Warfarin was stopped 2 days prior to
surgery, but Heparin infusion continued, keeping activated partial
thromboplastin time (APTT) twice the norm. At operation a right
Video assisted thoracoscopic procedure (VATS) was performed. The
chest was found to be full of thin fluidy amber colour pus. The superior
mediastinum was bulging especially between superior vena cava and
main trachea, as well as anterior to the SVC, and below Azygos vein in
the subcarinal space deep into the posterior mediastinum.
On opening the pleural reflection on these sites, frank yellow
projectile thick yellow pus exuded (Figure 3). The lung parenchyma
at the back of the hilum was exuding pus from metastatic lung
abscesses. The pericardium anterior to the phrenic nerve was also
bulging and tense, and was therefore aspirated with a needle and
blood stained fluid was obtained, therefore; was not opened to avoid
purulent pericarditis. The pleural reflection over the superior triangle (between vagus, phrenic nerve and Azygos vein) was opened and
pus drained (Figure 3A, 3B, 3C). Likewise the subcarinal space was
opened longitudinally between Azygos vein and oesophagus down to
diaphragm (Figure 4A, 4B, 4C). The space anterior to the SVC was
opened and bluntly dissected under the left innominate vein across
the midline, up to the left side of neck and pus was drained. The
superior triangle and subcarinal spaces were joined below the Azygos
vein. Formal decortications of the rind on the visceral pleura was
carried out. This was not difficult as this was an early phase empyema.
The pleural cavity was washed with warm water and aqueous Iodine
solution mixed with 3% Hydrogen Peroxide (H2O2). A 32F drain with
4 extra holes was introduced via the inferior port and placed within
the mediastinum behind vena Azygos up to the superior triangle
(Figure 4C).
The patient was returned to ICU for respiratory and cardiac
support. A per-cutaneous tracheostomy was fashioned the following
day. After two weeks in ITU he was extubated. It appeared that he had
right sided weakness and expressive dysphasia. CT head was normal
but MRI showed 2 subcortical watershed infarcts. Aspirin was added
to his subcutaneous heparin. He was fed thorough nasogastric tube
and later via gastrostomy (PEG). A speech video fluoroscopy was
performed to elucidate whether the dysphagia was due to the stroke or
other causes. Oropharyngeal dysmotility with silent nasopharyngeal
and laryngeal penetration was reported, as well as minor aspiration
into trachea (Figure 2D). He made slow recovery and at the time of his
discharge he was mobilizing on the ward with satisfactory CT chest
and chest X-ray. He was discharge to a local hospital after 8 weeks
in our hospital. He was reviewed in the clinic 4 months later when
swallowing was back to normal and he could walk 5 miles without
stopping. His speech was still slightly tongue tied with some words.
Warfarin treatment was planned for 9 months.
Figure 1
Figure 1
Radiological images.
A: CT of Case No. 4 before right thoracotomy. 1= Gas and brawny oedema of
the mediastinum 2= left pleural chest drain, not draining effusion 3= bilateral
pleural effusions 4= patchy consolidation of the left lung.
B: CT of same patient one week after right thoracotomy. 1= Pre-Aortic
persistent gas and brawny oedema/pus of the mediastinum 2= left pleural
chest drain, not draining 3= Posterior Left pleural effusion 4= 36F mediastinal
drain between trachea and oesophagus 5= 32F drain in free right pleural
space.
C: Type I DNM in patient No. 2. Brawny oedema in neck and superior
mediastinum, not seen below level of carina.
D: Speech Video Flouroscopy in case No. 3. Oropharyngeal dysmotility, with
contrast crossing the vocal cords into the trachea (arrow).
Figure 2
Figure 2
VATS debridement of superior mediastinum.
A: Normal looking superior triangle (borders: phrenic nerve, vagus nerve
and Azygos vein). Dotted line represents where pleural reflection should be
incised.
B: Same view in Case 5 showing tense mediastinum bulging with abcesses,
before incision.
C: Same patient after incision debirdement and drainage. T= Trachea, SVC=
Superior Vena Cava.
Figure 3
Figure 3
VATS debridement of posterior mediastinum.
A: Bulging tense pleural reflection on posterior mediastinum below the level
of carina. VB= Vertebral Bodies.
B: Posterior mediastinum opened from concavity of Azygos to diaphragm.
C: A 36F Argyll drain with extra holes inserted under Azygos into superior
triangle, as high as thoracic inlet.
Discussion
DNM has been classified according to extension of the
inflammatory tracts into the mediastinum [6]. Type I lesions remain
above the carina, whereas type IIA lesions reach the anterior inferior
mediastinum, and type IIB lesions involve the posterior inferior
mediastimun. DNM has high mortality of 31% even with appropriate
treatment [7]. Delay in diagnosis and inadequate drainage are thought
to be the cause for this high mortality. When infection spreads below
the level of the carina, mediastinal drainage should be performed
through a thoracic approach, as mortality associated with cervical
drainage alone is 47% compared to 19 % for combined cervical and
thoracic drainage [7]. There is no widely adopted consensus as to the
best surgical approach. Ridder et al. [8] discussed the contemporary
trends in management recently. Bilateral thoracotomy, subxiphoidal,
clamshell and median sternotomy amongst other approaches have
been described [7-13]. Full postero-lateral thoracotomy has been
traditionally advocated as it offers the best access to the prevertebral
and paraoesophageal planes without the risk of cutting through bone.
Sternotomy carries the risk of sternal breakdown and osteomyelitis.
In this selected series of five patients with DNM who were
managed at Southampton General Hospital – UK in the last 10
years, bespoke surgical approach was the theme. Table 1 summarises DNM has been classified according to extension of the
inflammatory tracts into the mediastinum [6]. Type I lesions remain
above the carina, whereas type IIA lesions reach the anterior inferior
mediastinum, and type IIB lesions involve the posterior inferior
mediastimun. DNM has high mortality of 31% even with appropriate
treatment [7]. Delay in diagnosis and inadequate drainage are thought
to be the cause for this high mortality. When infection spreads below
the level of the carina, mediastinal drainage should be performed
through a thoracic approach, as mortality associated with cervical
drainage alone is 47% compared to 19 % for combined cervical and
thoracic drainage [7]. There is no widely adopted consensus as to the
best surgical approach. Ridder et al. [8] discussed the contemporary
trends in management recently. Bilateral thoracotomy, subxiphoidal,
clamshell and median sternotomy amongst other approaches have
been described [7-13]. Full postero-lateral thoracotomy has been
traditionally advocated as it offers the best access to the prevertebral
and paraoesophageal planes without the risk of cutting through bone.
Sternotomy carries the risk of sternal breakdown and osteomyelitis.
In this selected series of five patients with DNM who were
managed at Southampton General Hospital – UK in the last 10
years, bespoke surgical approach was the theme. Table 1 summarises the demographics and outcomes of these patients. Our institution
previously published the importance of contrast CT imaging in
identifying the extent of infection and the presence of complications
[14]. Clinicians are reminded that the disease is not uniform, different
patients have different maps of pus locations and that mediastinal
drainage was the single most important factor in recovery. One should
not be under the illusion that bilateral large sized intercostals drains
would sort out the sepsis. Intact mediastinal pleural reflections are an
impediment to drainage of pus originating within the mediastinum.
Opening the mediastinal pleural reflections from neck to diaphragm
and entering into the loculations of pus is mandatory. Nakamori et al.
[15] seem to suggest that the principal objective of surgery is drainage
and not debridement. They successfully treated 6 patients with per
cutaneous mediastinal drainage. We agree with this principle as it
forms the basis for the success of minimal access surgery. However;
drainage must be accompanied by opening all loculations of pus for
proper drainage, and this might not amount to debridement.
In case No. 1, the J shaped manubriotomy offered direct access to
the anterior and superior mediastinum, especially the pre-aortic and
SVC spaces thus reducing surgical trauma. Both sterno-manubrial
joints were preserved. However; Saute et al. [16] stated that “sawing
healthy bone in the presence of infection is not a correct surgical
practice”. With improved VATS experience and if the patient
condition allows lung isolation, a right VATS approach will give a
similar exposure, if not better.
In case No. 2 there was a huge discrepancy between the CT
findings in the mediastinum and the clinical behavior. The CT
showed gross involvement of the superior mediastinum without
gross extension into the posterior mediastinum below level of
carina. The clinical response to antibiotics and the early spontaneous
discharge of the abscess into the mouth, necessitated a wait and watch
approach. Similar approach has been advocated recently by lyer et
al. and Jayasekera et al. [17,18]. The clinical improvement avoided
any need for surgical intervention, and this was a pleasant reminder
that clinicians treat patients and not their X-rays. Within two weeks
in hospital the symptoms abated and dysphagia resolved completely
without restricting oral intake. The radiological improvement lagged
roughly two weeks behind the clinical improvement. The other
point about the CT is its inability to differentiate between brawny
oedema and frank pus, especially when contrast is not given. The
former signifies thick indurated non-fluid necrosis that does not
require drainage, and if incised does not drain fluid pus, leading to
persistence of CT findings. Resolution of mediastinal brawny oedema
takes weeks long after the patient has turned the corner. Positron
Emission Tomography (PET) has been used in similar situations in
the past to reveal the site of “pus somewhere” [19].
In case No. 3 adequate drainage was achieved via a cervical
approach alone. The blind insertion of a large bore chest drain from
the neck parallel to the oesophagus all the way down to diaphragm
achieved adequate drainage. This emphasises the previously
mentioned principle of drainage versus debridement pointed out by
Nakamori et al. [15].
Case No. 4 represents the classical approach of bilateral full
thoracotomies, which surgeons might have to resort to, or will bail
out those with limited experience in VATS mediastinal surgery.
VATS can be difficult in many cases, but using CO2 insufflation can
help at times and low tidal volume, although VATS was decided in
the first instance in this case but due to sepsis, hypoxia and retention
of CO2 necessitated high inflation pressures making lung isolation
impossible therefore patient underwent thoracotomy. Drainage
by VATS on the left side can prove to be difficult with the greater
vessels in your way. However, VATS procedure on the right is a very
common procedure, especially when done in a VATS lobectomy for
cancer. Mediastinal lymph node dissections are done routinely and
a drainage procedure on the right would be a very simple procedure
indeed. Most of the times, if discovered early, a drain may be placed
in the neck and prevent further extension to the mediastinum
although a drain was placed early in the neck but due to extension
in the mediastinum warrant a surgical exploration in this case. We
believe that open procedure is not the answer and may only prove
to be a dangerous procedure considering it may only aggravate the
situation but if VATS cannot be performed either due to limited
experience or clinical condition than open procedure should be
considered. The mortality rates for mediastinitis are high (60%) [20]
And thus all options must be considered.
Case No. 5 reflects our experience with VATS at Southampton
General Hospital. We amassed a good experience with mediastinal
anatomy during VATS Systematic Nodal Dissection for lung cancer.
This proved very useful and the experience rolled over in extensive
mediastinal drainage, avoiding caveats and areas of danger. The
authors believe that a right sided VATS procedure is more than
capable of opening the mediastinum from neck to diaphragm,
making an additional left VATS unnecessary. Similar experience was
previously reported by Min et al. and Hsin et al. [21,22]. However;
patients could present in extremis, requiring high percentage of
Oxygen entraining and high inflation pressures as in case No. 4
making VATS not an option in such cases. Access to the left pleural
space for drainage of left effusion could be achieved from the right
side by retrosternal dissection and opening the contralaterl pleura
in the anterior mediastinum, again from neck to diaphragm. Dajer-
Fadel et al. [12] described a way of opening the left pleura from a right
thoracotomy approach between the Azygos vein and oesophagus just
above the diaphragm. Such an approach could be tried by VATS if
the necrosis allows a degree of anatomical identification of structures.
Mobilisation of the oesophagus in extreme necrotic conditions carries
the risk of perforation and injury to the thoracic duct. The weak point
in a right sided VATS approach is that it does not provide good access
to the pre-aortic space, at the level of the aortic arch in the left chest.
A left Chamberlain, a left VATS or a cervical midline approach might
complement a right VATS approach in such cases. Deep knowledge
of anatomical structures is required as major vessels are difficult to
identify within the debris of a purulent milieu. If the pockets of pus
are not all opened and drained, relapse of sepsis is almost guaranteed.
The use of large calibre drains such as 32F and 36F connected to
high suction of -3 to -5 Kilo Pascal ensures good drainage of necrotic
material as well as fluid pus. In our experience the use of mixed
Aqueous Iodine and H2O2 for wash out of the chest cavity is safe and
has the advantage of combating anaerobic organisms.
Recent improvement in ITU capabilities plays an important part in
the recovery of these patients, involving several teams of anaesthetist,
intensivists, radiologists, cardiologist, gastroenterologists, ENT,
maxillofacial and Thoracic surgeons, clinical microbiologists,
pharmacists, physiotherapists, nutritionists amongst other specialist
teams. The integrated care of all these teams and marginal gains
underpin a successful outcome. It is of note that none of the patients
included in this series were immune compromised, and were
previously fit and healthy. Three patients in this series were gravely ill, on mechanical ventilatory support for weeks, however; Acute
Respiratory Distress Syndrome (ARDS) or acute lung injury was
never a problem. Because of early intervention; late decortication
of the lung was never required. Exacerbation of febrile illness, rising
inflammatory markers and inotropic requirement were common
despite a good surgical drainage procedure. Unexpected turn to the
worse was an indication for emergency review by contrast CT scan,
from base of skull to abdomen. Usually a fresh collection of pus is
revealed, and multiple surgeries are required.
The striking symptom of dysphagia was common to all, albeit in
varying degrees, but was difficult to interpret. The pathophysiology
for dysphagia in DNM is ill understood, however numerous factor
interplay, e.g. pharyngeal sepsis (tonsillitis, Quincy, pharyngitis),
pharyngeal perforation, peri-oesophageal sepsis, external compression
by pus or enlarged nodes, drug induced and bacterial toxins induced
neurotoxicity to vagal nerve branches and the local Auerbach and
Meissners plexuses. Scientific investigation of this phenomenon
is not always possible as these patients could be debilitated and
mechanically ventilated in the ITU. Dysphagia is recognised when
patient resumes self-ventilation in the ITU. It has to be presumed in
all patients and SALT should be involved early, as it could lead to
regurgitation, aspiration and lung parenchymal infections including
abscess formation, further complicating the situation and adding
significantly to morbidity and mortality.
In conclusion, DNM is a rare but life-threatening complication
likely to be encountered once every 1-2 years by a UK thoracic surgeon.
It has a wide spectrum of presentations which requires targeted
individual management ranging from conservative management to
several surgical cervical and thoracic drainage procedures, guided
by contrast CT scans. These cases require multitude of specialised
teams and should be managed in a tertiary referral hospital. Minimal
access surgery (VATS) can play an important role and can achieve
the same objective as thoracotomy. Dysphagia is an ill understood
complication of this life-threatening sepsis and takes a long time to
recover.
Table 1
References
- Novellas S, Kechabtia K, Chevallier P, Sedat J, Bruneton JN. Descending necrotizing mediastinitis: a rare pathology to keep in mind. Clin Imaging. 2005;29(2):138-40.
- Bhattacharya S, Livsey SA, Wiselka M, Bukhari SS. Fusobacteriosis presenting as community acquired pneumonia. J Infect. 2005;50(3):236-9.
- Perez A, Cueto G, de la Escosura G, Cicero R. Descending Necrotizing Mediastinitis. Results of Medical-Surgical Treatment in 17 Cases. Gac Med Mex. 2003;139(3):199-204.
- Makeieff M, Gresillon N, Berthet JP, Garrel R, Crampette L, Marty-Ane C, et al. Management of descending necrotizing mediastinitis. Laryngoscope. 2004;114(4):772-5.
- Grodinsky M, Holyoke EA. The fasciae and fascial spaces of the head, neck and adjacent regions. Am J Anat 1938;63:367.
- Hasegawa T, Endo S, Sohara Y. Classification of descending necrotizing mediastinitis. Ann Thorac Surg. 2000;69(4):1296.
- Corsten MJ, Shamji FM, Odell PF, Frederico JA, Laframboise GG, Reid KR, et al. Optimal treatment of descending necrotising mediastinitis. Thorax. 1997;52(8):702-8.
- Ridder GJ, Maier W, Kinzer S, Teszler CB, Boedeker CC, Pfeiffer J. Descending necrotizing mediastinitis: contemporary trends in etiology, diagnosis, management, and outcome. Ann Surg. 2010;251(3):528-34.
- Marty-Ané CH, Berthet JP, Alric P, Pegis JD, Rouvière P, Mary H. Management of descending necrotizing mediastinitis: an aggressive treatment for an aggressive disease. Ann Thorac Surg. 1999;68(1):212-7.
- Papalia E, Rena O, Oliaro A, Cavallo A, Giobbe R, Casadio C, et al. Descending necrotizing mediastinitis: surgical management. Eur J Cardiothorac Surg. 2001;20(4):739-42.
- Freeman RK, Vallières E, Verrier ED, Karmy-Jones R, Wood DE. Descending necrotizing mediastinitis: An analysis of the effects of serial surgical debridement on patient mortality. J Thorac Cardiovasc Surg. 2000;119(2):260-7.
- Dajer-Fadel WL, Ibarra-Pérez C, Argüero-Sánchez R. Mediastinal drainage in descending necrotizing mediastinitis. Asian Cardiovasc Thorac Ann. 2013;21(4):493-5.
- Erkmen CP, Wang H, Czum J, Paydarfar JA. Use of cervicothoracic anatomy as a guide for directed drainage of descending necrotizing mediastinitis. Ann Thorac Surg. 2012;93(4):1293-4.
- Lyle NJ, Rutherford EE, Batty VB. A pain in the neck--imaging in neck sepsis. Clin Radiol. 2011;66(9):876-85.
- Nakamori Y, Fujimi S, Ogura H, Kuwagata Y, Tanaka H, Shimazu T, et al. “Conventional open surgery versus percutaneous catheter drainage in the treatment of cervical necrotizing fasciitis and descending necrotizing mediastinitis,” AJR Am Roentgenol. 2004;182:1443–49.
- Saute M. Descending necrotizing mediastinitis: an old issue with an new approach. Eur J Cardiothorac Surg. 2012; 42:e73.
- Iyer S, Collum J, Babores M. Descending necrotizing mediastinitis: a conservative approach. Ear Nose Throat J. 2014;93(3):E11-4.
- Jayasekera BA, Dale OT, Corbridge RC. Descending necrotising mediastinitis: a case report illustrating a trend in conservative management. Case Rep Otolaryngol. 2012;2012:504219.
- Mandry D, Tatopoulos A, Chevalier-Mathias E, Lemarié J, Bollaert PE, Roch V, et al. ¹8F-fluorodeoxyglucose positron emission tomography combined with whole-body computed tomographic angiography in critically ill patients with suspected severe sepsis with no definite diagnosis. Eur J Nucl Med Mol Imaging. 2014;41(10):1924-30.
- Erkmen CP, Wang H, Czum J, Paydarfar JA. Use of cervicothoracic anatomy as a guide for directed drainage of descending necrotizing mediastinitis. Ann Thorac Surg. 2012;93(4):1293-4.
- Min HK, Choi YS, Shim YM, Sohn YI, Kim J. Descending necrotizing mediastinitis: a minimally invasive approach using video-assisted thoracoscopic surgery. Ann Thorac Surg. 2004;77(1):306-10.
- Hsin MK, Yim AP. Video assisted thoracoscopic surgery is a valuable approach for the management of descending necrotizing mediastinitis. Ann Surg. 2011;253:1055.