ITMDDF

ecc

ft5tsifira

upt

Tcoiplcrb

ttofdwscs

A

AHa

©P

GEN

ERA

LT

HO

RA

CIC

ndications and Results of Anterior Mediastinalracheostomy for Malignanciesassimo Conti, MD, Lotfi Benhamed, MD, Geoffrey Mortuaire, MD,ominique Chevalier, MD, Claire Pinçon, PhD, and Alain Wurtz, MD

epartments of Thoracic Surgery and Otolaryngology, Lille University Teaching Hospital, Lille; University of Lille - North of

rance, Lille; and Department of Biostatistics, UDSL, Lille, France

oeptcmgsp5

lmglwc

Background. This study was designed to assess thearly and long-term results, and determine suitable indi-ations of anterior mediastinal tracheostomy, after radi-al resection for cervicothoracic malignancies.

Methods. From 1985 to 2009, 13 patients ranging in agerom 40 to 75 years underwent an anterior mediastinalracheostomy for malignancy, 2 as an isolated procedure,

with concomitant laryngectomy, and 6 with concomi-ant laryngopharyngoesophagectomy. The patients hadubglottic carcinoma, proximal tracheal carcinoma invad-ng the subglottic larynx (n � 6), stomal recurrenceollowing laryngectomy (n � 4), esophageal carcinomanvading the proximal trachea (n � 2), or tracheal recur-ence after conventional resection (n � 1). The postoper-tive course and outcome were assessed in all patients.Results. In 5 patients, the postoperative course was

neventful. There were 2 in-hospital deaths, and 6atients experienced nonfatal complications including

hree pharyngocutaneous fistulas, two pharyngogastric

P

PF1t1(cisr

ahus

raitlc

pôpital Albert Calmette, CHU Lille, F59037 Lille CEDEX France; e-mail:

lain.wurtz@chru-lille.fr.

2010 by The Society of Thoracic Surgeonsublished by Elsevier Inc

r pharyngocolic anastomotic leaks, and one pulmonarymbolism. Satisfactory airway was achieved in 7 patientsresenting with proximal obstruction, and ability to

olerate oral feeding, in 2 patients with esophageal car-inoma. The mean follow-up time for survivors was 89onths (range, 9 to 201 months). Patients with esopha-

eal carcinoma or recurrence of tracheal carcinomahowed a poor outcome. In contrast, in the subgroup ofatients with head and neck malignancy, 3-year and-year survival rates were 57% and 43%, respectively.Conclusions. Our results and survival analysis from the

iterature suggest that suitable indications for anteriorediastinal tracheostomy are (1) carcinoma of the sub-

lottic region or proximal trachea invading the subglotticarynx, (2) stomal recurrence after laryngectomy, and (3)ell-differentiated thyroid carcinoma invading the tra-

hea or recurrence.(Ann Thorac Surg 2010;89:1588–96)

© 2010 by The Society of Thoracic Surgeons

he surgical treatment of extensive cervicothoracicmalignancies requires a laryngotrachiectomy or, in

ase of stomal recurrence after laryngectomy, a divisionf the retrosternal trachea. Furthermore, when major

nvasion of the proximal esophagus has occurred, com-lete exenteration (laryngopharyngoesophagectomy) fol-

owed by esophageal substitution is required. In mostases, the adequate, 15-mm-long tracheal margin [1]equires the division of the retrosternal trachea followedy establishment of a mediastinal stoma.Given the rarity of the anterior mediastinal tracheos-

omy (AMT), there are only sparse series including morehan 10 patients in the English literature [2–7], and somef them lack details about the patients’ outcome. There-ore, to assess the early and long-term results, and toetermine suitable indications of AMT for malignancy,e reviewed the postoperative course and outcome in a

eries of 13 consecutive patients who underwent thishallenging procedure in our institution, drawing conclu-ions from pooled evidence from other series.

ccepted for publication Jan 21, 2010.

ddress correspondence to Dr Wurtz, Pôle de Chirurgie Thoracique,

atients and Methods

atientsrom July 1985 to January 2009 we performed 15 AMTs,3 for malignancy and 2 for nonmalignant tracheodiges-ive fistula. Among the first group of patients, there were2 men and 1 woman ranging in age from 40 to 75 yearsmean, 56 � 11 years; Table 1). The present study in-ludes a prior report [8] and received approval from ournstitutional review board, with waiver of informed con-ent, according to French regulations applying to suchetrospective studies.

Of these 13 patients, 6 had a prior history of surgerynd postoperative radiotherapy between 50 and 55 G foread and neck malignancy. Additionally 1 of these 6 hadndergone a right lower lobectomy for a synchronousquamous cell bronchial carcinoma.

Six patients had primary carcinoma of the subglotticegion or proximal trachea invading the subglottic larynx,nd 2, squamous cell carcinoma of the esophagus, in whichnvasion of the proximal trachea had occurred. Four pa-ients were found to have type 1 or 2 stomal recurrence afteraryngectomy [9], and 1, tracheal anastomotic relapse afteronventional resection and postoperative radiotherapy.

Owing to the severity of their airway obstruction, 2

atients with primary carcinoma of the subglottic region

0003-4975/10/$36.00doi:10.1016/j.athoracsur.2010.01.044

Table 1. Clinical Characteristics, Procedures, Pathologic Findings, and Outcomes in Patients Undergoing Anterior Mediastinal Tracheostomy for Malignancy

PatientNo. Sex

Age(y) Malignancy History Diagnosis Procedure

PostoperativeComplications

PathologicFindings

Curative orPalliativeResection Outcome

TrachealMargin

NodalStatus

1 F 75 TL, RT Stomal recurrence AMT R0 3N� CR Dead, 70 mo2 M 48 TL, RT Stomal recurrence AMT, TR R1 N� PR Dead, 16 mo3 M 53 None Tracheal SCC invading the

subglottic larynxTL, AMT, TR R0 1N� CR Dead, 201 mo

4 M 53 None Esophageal SCC CE, AMT, TR, gastricpull-up

R0 4N� PR Dead, 6 mo

5 F 63 None Esophageal SCC CE, AMT, TR, colonicinterposition

Anastomotic leak R1 2N� PR Dead, 4 mo

6 M 59 Pelvimandibulectomy,RT

Subglottic SCC CE, AMT, TR, gastricpull-up

Flap failure R0 N� CR Postoperative death

7 M 47 TL, RT, lung lobectomy Stomal recurrence CE, AMT, gastricpull-up

R1 N� PR Dead, 20 mo

8 M 53 None Subglottic SCC TL, AMT, TR PC fistula R1 5N� PR Dead, 57 mo9 M 53 PL, RT Subglottic SCC (recurrence) TL, AMT PC fistula R0 N� CR Postoperative death

10 M 75 TL, RT Stomal recurrence CE, AMT, TR, gastricpull-up

Anastomotic leak R0 N� CR Alive, 90 mo follow-up

11 M 40 None Subglottic ACC TL, AMT, TR PC fistula R0 1N� CR Alive, 56 mo follow-up12 M 65 Tracheal resection, RT Anastomotic tracheal relapse CE, AMT, TR, gastric

pull-upPulmonary

embolismR0 5N� PR Dead, 11 mo

13 M 42 None Subglottic ACC TL, AMT, TR PC fistula R0 2N� CR Alive, 9 mo follow-up

ACC � adenoid cystic carcinoma; AMT � anterior mediastinal tracheostomy; CE � cervical exenteration; CR � curative resection; F � female; M � male; N� � no nodalmetastasis; N� � nodal metastasis; PC fistula � pharyngocutaneous fistula; PL � partial laryngectomy; PR � palliative resection; R0 � tracheal margin negative; R1 � tracheal marginpositive; RT � radiation therapy; SCC � squamous cell carcinoma; TL � total laryngectomy; TR � tracheal relocation.

1589A

nn

Th

oracSu

rgC

ON

TI

ET

AL

2010;89:1588–96

AN

TE

RIO

RM

ED

IAST

INA

LT

RA

CH

EO

STO

MY

GENERAL THORACIC

u2

tacpensubuttppsu

t

OIua

E

tahsm

tifssr

wupettloj

ptfi

C

taamvasiin

Fbtn

1590 CONTI ET AL Ann Thorac SurgANTERIOR MEDIASTINAL TRACHEOSTOMY 2010;89:1588–96

GEN

ERA

LT

HO

RA

CIC

nderwent emergency Nd-YAG-laser deobstruction, and, transtumoral tracheostomy before considering AMT.All patients underwent thorough preoperative func-

ional and oncologic assessment, including physical ex-mination, performance status, pulmonary function tests,ervical, chest and upper abdominal computed tomogra-hy, and, in both patients with esophageal carcinoma,sophagography and esophagoscopy. Fiberoptic exami-ation of the pharyngolarynx, trachea, and bronchi wasystematically performed, followed by panendoscopynder general anesthesia for proper histologic diagnosticiopsy, if required. Patient 10 underwent an esophagealltrasonography because of close contact between a large

racheostomal recurrence and his esophagus to assesshe esophageal muscular layer extension. Recently,ositron emission tomography was performed in the lastatient with subglottic adenoid cystic carcinoma andhowed no significant avidity in the cervical area andpper mediastinum.In all patients, the distal trachea was macroscopically

umor free for the last 5 cm up to the carina.

perative Techniquen all patients, the mediastinal stoma was established bysing a pectoralis major myocutaneous island flap (PMF),s previously described [8–10].

XPOSURE AND TUMOR RESECTION. All patients were posi-ioned supine, and the approach was a combined collarnd median incision, with manubrium and clavicularead removal performed only when cervical tumor dis-ection [4] or frozen section analysis of the trachealargin determined the need for a retrosternal division of

ig 1. Operative view of the tracheal stumpefore (A) and after (B) relocation inferior tohe innominate artery and up to the left in-ominate vein.

he trachea. This approach enabled laryngotrachiectomyn 5 patients, or elliptical skin incision around the stomaollowed by division of the retrosternal trachea for type 1tomal recurrence in 2 patients. Neck dissection anduperior mediastinal lymphadenectomy completed theesection step.

Two patients with esophageal carcinoma and 4 inhom major invasion of the esophagus had occurrednderwent a complete exenteration. The transhiatal ap-roach was used to perform a blunt transmediastinalsophageal dissection and combined resection of theumor, including the entire laryngopharynx, proximalrachea, and esophagus. Gastric pull-up in 5 patients oreft colonic interposition in 1 achieved the reconstructionf the upper digestive tract to the pharynx. Feeding

ejunostomy was routinely performed.In either standard AMT or AMT associated with com-

lete exenteration, removal of the tumor required ahyroid lobectomy in 5 patients, or a total thyroidectomyollowed by systematic parathyroid gland reimplantationn 4 (patients 4, 5, 10, and 11).

ONSTRUCTION OF THE MEDIASTINAL STOMA. After completeumor resection ascertained by systematic frozen-sectionnalysis of the tracheal margin, the PMF based on itscromiothoracic blood supply, with the entire pectoralisajor muscle and an oval-shaped skin island, was ele-

ated for reconstruction of the cervical and upper medi-stinal defects. Because the residual trachea was toohort to perform a standard stoma, it was relocalizednferior to the innominate artery [11, 12] and up to the leftnnominate vein in 10 patients (Fig 1). The left innomi-ate vein was not divided in all but 2 patients, in whom

t3emgaliorwFlK

2a

PIthaspic

gnacssH

pdsartld

wtli(

FTodllesdsf

FmTm

F

1591Ann Thorac Surg CONTI ET AL2010;89:1588–96 ANTERIOR MEDIASTINAL TRACHEOSTOMY

GEN

ERA

LT

HO

RA

CIC

he residual trachea was less than 5 cm (minimal length,.4 cm) after additional tracheal margin resection. Inither method, the underlying portion of the pectoraluscle created an interposition between the trachea and

reat vessels, and the skin island (6 to 7 cm) provideddditional coverage above the stoma, reducing sutureine tension between the inferior margin of the skinsland to the tracheal membrane (Fig 2). The tracheal ringf the stoma was sutured to the skin overlying theemaining sternum inferiorly, and the chest donor defectas closed primarily by mobilizing the lateral chest flaps.inally, the mediastinal stoma was cannulated with aow-pressure cuff cannula (Tracheoflex; Willy-Rüsh,ernen, Germany).The mean procedure duration was 456 minutes (range,

70 to 751 minutes). The patients’ clinical characteristicsnd procedures are summarized in Table 1.

ostoperative Caren patients undergoing cervical exenteration, postopera-ive mechanical ventilation during an average time of 12ours was required, owing to the length of the operationnd combined cervical, thoracic, and abdominal inci-ions. Constant humidification through nebulization wasrovided to the mediastinal stoma, and tracheal suction-

ng was carefully carried out to avoid obstruction, asso-iated with on-demand fiberoptic bronchoscopy. Naso-

ig 2. Operative view of reconstruction of the cervical and upperediastinal defects by using a pectoralis major myocutaneous flap.he inferior margin of the skin island was sutured to the tracheal

oembrane.

astric or jejunostomy tubes ensured satisfactory enteralutrition, until pharyngeal suture or pharyngodigestivenastomosis healing. Until tracheostomal healing, theannula was changed over a guide, by using either ahortened plastic tube (Biesalski; Willy-Rüsh; Fig 3) or ailicone tube (PROVOX Lary Tube; ATOS Medical,örby, Sweden), in the last 3 patients.To assess the early results, the postoperative com-

lications were classified as medical (pulmonary, car-iovascular complications, hepatic or renal failure,ystemic infection, and permanent iatrogenic hypopar-thyroidism) or related to surgical procedure (arterialupture, flap failure, tracheal necrosis, stomal disrup-ion, pharyngocutaneous fistula, and anastomoticeaks). Operative mortality included patients who dieduring hospitalization.Tumoral resection was deemed as potentially curativehen no microscopic tumor was found at the level of the

racheal margin (R0), and with a number of metastaticymph nodes (�4). It was deemed as potentially palliativen patients with positive tracheal margin (R1) or multiple�4) metastatic lymph nodes.

ollow-Up and Statistical Analysishe survival time was defined as the time from the datef the operation to the date of death attributable toisease. The patients who were alive at the date of the

ast follow-up were censored. The Kaplan-Meier productimit method was used to plot survival curves and tostimate the 3- and 5-year survival rates and the medianurvival times. Given the small sample size, no confi-ence interval was computed for the 3- and 5-yearurvival rates; when available, 95% confidence intervalsor the median survival times were constructed by means

ig 3. Postoperative image with stomal cannulation in patient 11.

f Brookmeyer and Crowley’s method [13].

S

R

PCpcwcwph(aa

ITrsdmtd1tcva

ahtgqflea

ts

bicsrlams

ug1f

AErst

LF

poAas

io1hd

sF

a(simyampcppmcr

(c4ttcr4m

C

Teciat3r

1592 CONTI ET AL Ann Thorac SurgANTERIOR MEDIASTINAL TRACHEOSTOMY 2010;89:1588–96

GEN

ERA

LT

HO

RA

CIC

The statistical analysis was performed, by using theAS system (SAS v8; SAS Institute, Cary, NC).

esults

athology Findingsomplete macroscopic resection was achieved in allatients. Of 13 resected tumors, there were 11 squamousell and 2 adenoid cystic carcinomas. Among the patientsith squamous cell carcinoma, resection was potentially

urative in 5 and palliative in 6. Four R1 patients, inhom frozen-section analyses showed high-grade dys-lasia at the level of the tracheal margin, were found toave in situ carcinoma (n � 3) or perineural invasion

n � 1) after final pathologic studies. In both patients withdenoid cystic carcinoma, resection was potentially cur-tive (Table 1).

n-Hospital Morbidity and Mortalityhere were 2 in-hospital deaths. Patient 6, in whom theight PMF had been used for prior head and neckurgery, underwent ligation of the left innominate veinuring the procedure. In fact, after reconstruction of theediastinal stoma by using the left PMF, venous conges-

ion and then flap necrosis occurred, leading to stomalisruption, innominate artery rupture, and death on day7. Patient 9 experienced a pharyngocutaneous fistulahat spontaneously resolved and died because of suddenardiac arrest on day 18. Post-mortem examination re-ealed invasive aspergillosis with numerous myocardialspergillosis foci.Six patients had nonfatal complications. Patients 8, 11,

nd 13 experienced a pharyngocutaneous fistula thatealed without reoperation. After cervical exentera-

ion, patients 5 and 8 had a pharyngocolic or pharyn-ogastric anastomotic leak, respectively, which re-uired rotation of the contralateral pectoral muscleap to successfully close the defect. Patient 12 experi-nced a pulmonary embolism, which resolved undernticoagulant treatment.Prolonged mechanical ventilation was required in pa-

ients 5, 6, and 10 for respiratory failure in a setting ofepsis arising from surgical complications.

All patients in whom the parathyroid glands hadeen reimplanted showed transient hypoparathyroid-

sm and became normocalcemic in the mid-termourse. However, patient 4 experienced bone metasta-is and died of hypercalcemia. Finally, because of prioradiotherapy and despite preservation of one thyroidobe, patient 12 experienced iatrogenic hypothyroidismnd hypoparathyroidism, which required thyroid hor-one replacement and oral calcium plus vitamin D

upplementations.Finally, in the present series, postoperative course was

neventful in 5 patients. In the remaining patients, sur-ical and medical complication rates were 46.2% and5.3%, respectively, and the length of hospital stay was

rom 12 to 101 days (mean, 29 � 23 days). o

djuvant Therapyight patients received a postoperative radiotherapy or aedo irradiation of between 31.5 and 56 G for lymphaticpread or positive tracheal margin. No adjuvant chemo-herapy was administered in any patient.

ong-Term ResultsUNCTIONAL RESULTS. Of 11 patients surviving surgery, 7resented with partial or complete subglottic or trachealbstruction, and satisfactory airway was achieved withMT. However, one of these, who refused to use a stomalppliance in the early postoperative period, experiencedtenosis, which required successful revision at 8 months.

Palliation of the esophageal obstruction was achievedn both patients who had esophageal carcinoma. Of thether 3 patients surviving cervical exenteration, patient0 who experienced a pharyngogastric anastomosis leakad an anastomotic stenosis, which required on-demandilatations, to achieve the ability to have oral feedings.Finally, all but 1 patient surviving surgery achieved

atisfactory esophageal or gastric speech.OLLOW-UP AND SURVIVAL. No patient was lost to follow-up,nd the overall mean follow-up time was 41.7 monthsrange, 1 to 201 months). Among the 11 patients survivingurgery, the resection was deemed as potentially curativen 6 and palliative in 5. In the first group, patient 1 died of

etastatic disease at 70 months, patient 3 survived 17.5ears and died of myocardial infarction, and patients 10nd 13 are still alive with no evidence of disease (90-onth and 9-month follow-up, respectively). Finally,

atient 11 is still alive with lung metastasis and receivinghemotherapy with 56 months’ follow-up. In contrast,atients with palliative resection were found to have aoor outcome and died of disease between 4 and 20onths, except for patient 8 with subglottic squamous

ell carcinoma who survived 57 months and died ofegional recurrence (Table 1).

The mean follow-up time for survivors was 89 monthsrange, 9 to 201 months). The Kaplan-Meier survivalurve shows that overall 3- and 5-year survival rates were3% and 32%, respectively (Fig 4); the median survivalime was 20 months (95% confidence interval, 6 to 70). Inhe subgroup of patients with head and neck malignan-ies (carcinomas of the subglottic region and stomalecurrence), 3- and 5-year survival rates were 57% and3%, respectively; the median survival time was 57onths (95% confidence interval not available).

omment

he construction of a mediastinal stoma after resection ofxtensive cervicothoracic malignancy is a major surgicalhallenge, which has been associated with high morbid-ty and mortality, particularly as a result of innominatertery rupture. In historical series before 1980, the mor-ality rate attributable to arterial rupture was between3% and 50% [2, 11, 14]. This devastating complicationesulted from tracheal stoma separation, with exposure

f the great vessels, mediastinal sepsis, pressure necrosis

ftclmio

ssaAtputaoa1ta

lcwubctmpnae

ptltctttpar

eslarte

lataofti

slmctalcmfriwpat

boTduw

hrcant

Fgpap

1593Ann Thorac Surg CONTI ET AL2010;89:1588–96 ANTERIOR MEDIASTINAL TRACHEOSTOMY

GEN

ERA

LT

HO

RA

CIC

rom the trachea to the innominate artery, or a combina-ion thereof. The key points to avoiding these majoromplications are to (1) preserve an adequate trachealength up to the carina, (2) fill the dead space in the upper

ediastinum, (3) avoid apposition of the trachea to thennominate artery, and (4) reduce skin suture line tensionf the stoma.The remaining tracheal length for the construction of a

afe AMT is controversial. In 1999, Orringer [15] empha-ized that “a 5-cm minimum length of remaining tracheabove the carina was optimal for the construction of anMT,” but recent literature focuses on the feasibility of

he procedure with a shorter tracheal stump [16]. In ouratients, the gross tumor extension did not exceed 5 cmp to the carina. When the residual trachea after addi-

ional tracheal margin resection was too short to performstandard stoma, we routinely performed the relocation

f the trachea inferior to the innominate artery [11, 12] tovoid a potentially hazardous ligation of this vessel [7,6]. Finally, in 2 patients with a residual trachea of lesshan 5 cm, AMT could be achieved, which needed andditional division of the left innominate vein.In the present study, the mediastinal stoma was estab-

ished by using a PMF, as described first by Biller andolleagues [10] in 1980. This confident flap providedell-vascularized muscle to fill the dead space in thepper mediastinum and created a bulky interpositionetween the trachea and the innominate artery. Theutaneous paddle of the flap reduced skin suture lineension, avoiding tracheal stoma separation or necrosis,

ainly in the challenging cases of patients who receivedrior radiotherapy for head and neck malignancy. Fi-ally, we observed one flap failure leading to innominatertery rupture as the consequence of a strategic error,arly in our experience.

ig 4. Kaplan-Meier curve for overall survival in 13 patients under-oing an anterior mediastinal tracheostomy for malignancy. Twoatients were censored after 70 months, with a follow-up time of 90nd 201 months. From 70 to 201 months, the estimated cumulativerobability of survival was identical to that at 70 months.

Pharyngocutaneous salivary fistula is a common com- r

lication in the postoperative period after total laryngec-omy, with a reported incidence up to 65%. Hemoglobinevel less than 12.5 g/dL, prior tracheostomy, preopera-ive radiotherapy, and concurrent neck dissection areurrently considered as significant risk factors [17]. Fur-hermore, concurrent mediastinal dissection weakenedhe pharyngoesophageal tube vascularization, which po-entially increased the risk of fistula. In fact, 4 of 5atients had two (n � 3) or three (n � 1) risk factors [17]nd experienced such a fistula that resolved withouteoperation [18].

Among the 4 patients who underwent a completexenteration followed by gastric pull-up and survivingurgery, 1 showed a pharyngogastric anastomotic leak, inine with the incidence observed by Orringer of 29% [5]nd Maipang and associates of 18% [7]. The patient whoeceived a colonic interposition experienced such anas-omotic complication as well. In both cases, a contralat-ral pectoral muscle flap successfully closed the defect.Fortunately, patients who had fistula or anastomotic

eak did not experience any mediastinitis or mediastinalbscess, demonstrating the efficiency of PMF to preventhese devastating complications. In a similar way, thedditional skin coverage prevented the need for numer-us stomal revisions in the mid-term course [4] andacilitated the postoperative radiotherapy, without mid-erm and long-term adverse effects, even in case of redorradiation [8].

Since 1980, there are sparse monocentric studies de-cribing the AMT postoperative course in the Englishiterature [4–7, 9, 10, 12, 19, 20] with only four reporting

ore than 10 patients (Table 2). Of these, one reportedervical exenteration with or without AMT reconstruc-ion for malignancy or postirradiation damage [4], andnother one, AMT for malignant or benign diseases [5],eading to potential discrepancies in data analyses. Inancer patients, the literature data showed an in-hospitalorbidity rate from 25% to 57%, and a mortality rate

rom 0% to 25%. Even though innominate artery ruptureemained the main cause of postoperative death, itsncidence dramatically decreased (5.3%) in comparisonith historical series [2, 11, 14], thanks to reconstructiverocedures using PMF [7, 9, 12, 19] or omental flapsssociated with innominate artery division [4] or reloca-ion [5, 20] (Table 2).

Long-term results are difficult to estimate as well,ecause some studies did not give any information [9, 10]r had a lack of details [2, 12, 20] about patients’ outcome.hus, to assess the potential benefits of AMT, we pooledata for patients with 3-year survival according to thenderlying disease process from the sparse series inhich long-term results were reported [4–7, 20] (Table 3).The patients with esophageal carcinoma were found to

ave a poor outcome (3-year survival, 11%), and chemo-adiotherapy associated with palliative procedures (tra-heal stent, esophageal dilatations) is a currently reliablelternative therapy. In the field of head and neck malig-ancies, the stomal recurrence after laryngectomy was

he first indication for AMT [3, 4, 9]. In contrast to poor

esults reported in historical series [11, 21], data analyses

dtrsaao

ar1rmv

tHmap

otib

cmo

TS

R

BS

NF

GO

K

M

KP

a

eu

A neous

T

R

GOKMKPT

a

A

1594 CONTI ET AL Ann Thorac SurgANTERIOR MEDIASTINAL TRACHEOSTOMY 2010;89:1588–96

GEN

ERA

LT

HO

RA

CIC

emonstrated a rather favorable prognosis after resec-ion, with a 3-year survival of 44%. On the other hand,esults of AMT in advanced pharyngeal and laryngealquamous cell carcinomas, or resection as salvage ther-py, were extremely discouraging [5], and the low prob-bility for a long survival does not justify this type ofperation.In the field of carcinomas of the subglottic region

nd upper trachea, Grillo and coworkers [4] reported 3elevant cases of adenoid cystic carcinoma with 48-,18-, and 120-month survival. In the present study, weeport 2 additional cases, notably a patient surviving 56onths, despite worse prognosis factors such as ad-

anced tumor, preoperative transtumoral tracheos-

able 2. In-Hospital Morbidity and Mortality After Anteriorince 1980

eferenceNo. of Patients WithAMT Reconstruction

MortNo. of Pa

(IA Ru

iller 1980 [10] 7 (PMF) . .isson 1981 [9] 8 (PMF) 1 (12

(IA rupturelfeld 1984 [12] 8 (PMF) . .

ujita 1990 [19] 8 (PMF, n � 5) 2 (25(IA ruptur

rillo 1990b [4] 14 (epiploon) 1 (7%rringer 1992c [5] 40 (epiploon) 6 (15

(IA rupturato 1994 [6] 16 (miscellaneous) 3 (18

(IA rupturaipang 1996 [7] 11 (PMF) 2 (18

(IA rupturuwabara 2001 [20] 7 (epiploon) . .resent study 2010 13 (PMF) 2 (15

(IA ruptur

Death as a result of IA rupture in patients who did not receive PMFxenteration including 4 patients who did not undergo AMT. c Morbindergoing AMT for benign disease.

MT � anterior mediastinal tracheostomy; PMF � pectoral myocuta

able 3. Three-Year Survival After Anterior Mediastinal Trac

eference

Esophageal SCC Stomal Recurren

No. ofPatients 3YS

No. ofPatients 3YS

rillo 1990 [4] 7 1 . . .rringer 1992 [5] 16 2 5 2ato 1994 [6] 16 0 . . .aipang 1996 [7] 6 1 . . .

uwabara 2001 [20] 7 2 . . .resent study 2010 2 0 4 2otal 54 6 (11%) 9 4 (44%

Two patients with no evidence of disease were lost to follow-up at 2 ye

CC � adenoid cystic carcinoma; SCC � squamous cell carcinoma; 3Y

omy, and lymph node spread with capsular rupture.owever, a recent report suggests that such a tumoray respond to a combination of carboplatin/paclitaxel

nd radiation therapy, allowing an organ-sparing ap-roach [22].In a similar way, the recent literature focuses attention

n the value of conservative laryngotracheoplastic resec-ion for primary tumors of the proximal airway [23], andn patients with a limited lesion, it should be discussedefore considering AMT.Finally, in patients with well-differentiated thyroid

arcinoma invading the trachea or recurrence, AMTight be considered as well because of a 3-year survival

f 44% (Table 3). However, as has been recently shown,

astinal Tracheostomy for Malignancy: Monocenter Series

(%))

MorbidityNo. of Patients (%)

Uneventful PostoperativeCourse (%)

2 (29%) 71%2 (25%) 62,5%

3 (37.5%) 62.5%

1)a3 (37.5%) 37.5%

10 (55.5%) Estimate, 39%

1)25 (57%) Estimate, 30%

2)a4 (25%) 56.2%

1)3 (27%) 55%

3 (43%) 57%

1)6 (46.2%) 38.5%

AMT reconstruction. b Morbidity in a series of 18 cases of cervicalnd uneventful postoperative course in a series of 44 patients including 4

island flap.

tomy According to the Underlying Disease

Advanced orRecurrent

Laryngeal orPharyngeal SCC

Subglottic andTracheal SCC or

ACC

Well-DifferentiatedThyroid

Carcinoma

No. ofPatients 3YS

No. ofPatients 3YS

No. ofPatients 3YS

2 0 3 2 3 215 0 . . . 2 1. . . . . . . . .. . . . . . 4a 1. . . . . . . . .

1 0 5 3 . . .20 0 8 5 (62%) 9 4 (44%)

Medi

alitytientspture

..5%)e).

%)e, n �

)%)e, n �

.8%)e, n �

%)e, n �

..3%)e, n �

fordity a

heos

ce

)

ars.

S � 3-year survival.

sm

lAasergnTp

t(ssln

W

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

I

CeodlTbfplt

1595Ann Thorac Surg CONTI ET AL2010;89:1588–96 ANTERIOR MEDIASTINAL TRACHEOSTOMY

©P

GEN

ERA

LT

HO

RA

CIC

egmental laryngotracheal resection or reconstructionight be discussed whenever possible [24, 25].Therefore, we believe that extensive resection of ma-

ignancies in the cervicothoracic region, followed byMT reconstruction, might be considered with accept-ble morbidity and mortality after appropriate patientelection, including performance status assessment andvaluation by a multidisciplinary team including a tho-acic surgeon, radiation oncologist, and medical oncolo-ist. Additionally, frozen-section analysis of the lymphodes is essential at the time of initial neck exploration.herefore, the AMT procedure might be discarded inatients showing multiple (�4) metastatic nodes.Our experience and survival analyses from the litera-

ure suggest that suitable indications for AMT should be1) squamous cell or adenoid cystic carcinomas of theubglottic region and proximal trachea invading theubglottic larynx, (2) type 1 or 2 stomal recurrence afteraryngectomy, and (3) well-differentiated thyroid carci-oma invading the trachea or recurrence.

e are grateful to E. Kipnis for revision of the manuscript.

eferences

1. Orringer MB. Anterior mediastinal tracheostomy. J ThoracCardiovasc Surg 1979;78:850–9.

2. Terz JJ, Wagman LD, King RE, Brown P, Neilfed JP, Law-rence W. Results of extended resection of tumors involvingthe cervical part of the trachea. Surg Gynecol Obstet 1980;151:491–6.

3. Krespi YP, Wurster CF, Sisson GA. Immediate reconstruc-tion after total laryngopharyngoesophagectomy and medi-astinal dissection. Laryngoscope 1985;95:156–61.

4. Grillo HC, Mathisen DJ. Cervical exenteration. Ann ThoracSurg 1990;49:401–9.

5. Orringer MB. Anterior mediastinal tracheostomy with andwithout cervical exenteration. Ann Thorac Surg 1992;54:628–37.

6. Kato H, Tachimori Y, Watanabe H, et al. Mediastinal trache-ostomy during esophagectomy for cervicothoracic esopha-geal carcinoma invading the proximal trachea. J Surg Oncol1994;55:78–83.

7. Maipang T, Singha S, Panjapiyakul C, TotemchokchyakamP. Mediastinal tracheostomy. Am J Surg 1996;171:581–6.

8. Wurtz A, Darras J, Sobecki L, Maetz E, Quandalle P. Surgery

for malignant lesions of the upper trachea. Report of a series

ime; 29-day postoperative length of stay) cases with

mo

1

2010 by The Society of Thoracic Surgeonsublished by Elsevier Inc

of six cases of mediastinal tracheostomy. Ann Chir1988;42:209–13.

9. Sisson G, Goldman ME. Pectoral myocutaneous island flapfor reconstruction of stomal recurrence. Arch Otolaryngol1981;107:446–9.

0. Biller HF, Krespi YP, Lawson W, Baek S-M. A one-stage flapreconstruction following resection for stomal recurrence.Otolaryngol Head Neck Surgery 1980;88:357–60.

1. Waddel WR, Cannon BA. Technic for subtotal excision of thetrachea and establishment of a sternal tracheostomy. AnnSurg 1959;149:1–8.

2. Nelfeld JP, Theogaraj SD, Mehrhof AI. Reconstruction aftermediastinal tracheostomy. Am J Surg 1984;148:505–8.

3. Brookmeyer R, Crowley JA. A confidence interval for themedian survival time. Biometrics 1982;38:29–41.

4. Sisson G, Straehley J, Johnson NE. Mediastinal dissection forrecurrent cancer after laryngectomy. Laryngoscope 1962;72:1064–77.

5. Orringer MB. Anterior mediastinal tracheostomy with andwithout cervical exenteration. Ann Thorac Surg 1999;67:591.

6. Gómez-Caro A, Gimferrer JM, Macchiarini P. Technique toavoid innominate artery ligation and perform an anteriormediastinal tracheostomy for residual trachea of less than 5cm. Ann Thorac Surg 2007;84:1777–9.

7. Paydarfar JA, Birkmeyer NJ. Complications in head and necksurgery. Arch Otolaryngol Head Neck Surg 2006;132:67–72.

8. Galli J, De Corso E, Volante M, Almadori G, Paludetti G.Postlaryngectomy pharyngocutaneous fistula: incidence,predisposing factors, and therapy. Otolaryngol Head NeckSurg 2005;133:689–94.

9. Fujita H, Kakegawa T, Yamana H, Shirouzu G, Minami T.Mediastinal tracheostomy using a pectoralis major myocu-taneous flap after resection of carcinoma of the esophagusinvolving the proximal part of the trachea. Surg GynecolObstet 1990;171:403–8.

0. Kuwabara Y, Sato A, Mitani M, et al. Use of omentum formediastinal tracheostomy after total laryngoesophagectomy.Ann Thorac Surg 2001;71:409–13.

1. Sisson GA, Bytell DE, Edison BD, Yeh S. Transsternal radicalneck dissection for control of stomal recurrences-end re-sults. Laryngoscope 1975;85:1504–10.

2. Haddad RI, Posner MR, Busse PM, et al. Chemoradiotherapyfor adenoid cystic carcinoma. Preliminary results of an organsparing approach. Am J Clin Oncol 2006;29:153–7.

3. Gaissert HA, Grillo HC, Shadmehr MB, et al. Laryngotra-cheoplastic resection for primary tumors of the proximalairway. J Thorac Cardiovasc Surg 2005;129:1006–9.

4. Yu P, Clayman GL, Walsh GL. Human tracheal reconstruc-tion with a composite radial forearm free flap and prosthesis.Ann Thorac Surg 2006;81:714–6.

5. Gaissert HA, Honings J, Grillo HC, et al. Segmental laryn-gotracheal and tracheal resection for invasive thyroid carci-

noma. Ann Thorac Surg 2007;83:1952–9.

NVITED COMMENTARY

onti and colleagues [1] from France report a 24-yearxperience accumulating 13 cases of radical resectionf cervicothoracic malignancies requiring anterior me-iastinal tracheostomy. This small number is nonethe-

ess large in the world’s cumulative surgical literature.he operative endeavor is substantial, with apprecia-le operative morbidity and mortality, but is per-

ormed in the face of progressive disease robbing theatient of speech, swallowing, airway, and eventually,

ife. These are rare, tough (average 7.6-hour operative

any technical and management nuances, all nicelyutlined by the authors.Conti and colleagues [1] teach us several points:

. More than 7-year follow-up demonstrated exten-sive resection for cure of subglottic carcinoma orstomal recurrences supplies 5-year survival of 43%and 57-month median survival, whereas 3-yearsurvival for locally advanced proximal esophagealsquamous cell carcinoma or tracheal failure of prior

resection is zero.

0003-4975/10/$36.00doi:10.1016/j.athoracsur.2010.02.062