M

GEN

ERA

LT

HO

RA

CIC

Thoracoscopic Primary Esophageal Repair inPatients With Boerhaave’s SyndromeJeong Su Cho, MD, Yeong Dae Kim, MD, Jong Won Kim, MD, Ho Seok I, MD, and

in Su Kim, MD

Department of Thoracic and Cardiovascular Surgery, Pusan National University Hospital, Busan, Republic of Korea

Background. Early diagnosis and appropriate treatmentare important for a good outcome in Boerhaave’s syn-drome. The results of recent studies suggest that primaryesophageal repair should be performed for perforations,and some authors suggest that there are benefits fromthoracoscopic surgery in cases that are diagnosed early.

Methods. From December 2004 to May 2010, 15 patientswith Boerhaave’s syndrome presented to our department;the medical records were reviewed retrospectively forpreoperative signs and symptoms, interval between per-foration and surgery, surgical methods, and outcomes oftreatment. The patients were divided into two groupsaccording to the surgical approach (thoracoscopy versusthoracotomy) to evaluate the outcomes of thoracoscopicsurgery in patients with Boerhaave’s syndrome.

Results. All patients were men, with a mean age of 53.1years, and all underwent primary esophageal repair.

Seven patients underwent a thoracoscopic approach

domini@pnu.edu. This study was supported for two years by PusanNational University Research Grant.

© 2011 by The Society of Thoracic SurgeonsPublished by Elsevier Inc

(group A) and eight patients had a thoracotomy (groupB). The mean interval between perforation and surgerywas 43.5 hours (group A) and 40.2 hours (group B) (p �0.487). The mean operative time was 3.7 hours (group A)and 5.3 hours (group B) (p � 0.005). Postoperative leakswere confirmed by esophagography in one patient ingroup A and in two patients in group B. There was nomortality in group A and one death postoperatively ingroup B.

Conclusions. The results of this study suggest thatthoracoscopic esophageal repair may be a good surgicalalternative in patients with Boerhaave’s syndrome whohave a relatively stable vital sign or mild inflammation,regardless of the time interval between perforation andsurgery.

(Ann Thorac Surg 2011;91:1552–5)

© 2011 by The Society of Thoracic Surgeons

Boerhaave’s syndrome is a full-thickness transmuralrupture of the esophagus; it is the most severe

abnormality associated with esophageal perforation andhas a mortality rate ranging from 20% to 30% [1–4]. Delayin the diagnosis of Boerhaave’s syndrome is commonbecause of its nonspecific symptoms, which are oftenmisdiagnosed as acute pancreatitis, myocardial infarc-tion, or peptic ulcer disease. A delay in the diagnosisoccurs in more than 50% of patients and results in highmortality [5]; especially when treatment is delayed be-yond 48 hours. Such delays can lead to fatal mediastinitisand multisystem organ failure [2]. Therefore, early diag-nosis and appropriate treatment based on the patient’scondition are important for good outcomes. The resultsof recent studies recommend primary esophageal repairfor perforations, regardless of the time interval betweenperforation and surgery [6, 7], and some investigatorshave suggested that the minimally invasive thoraco-scopic technique is particularly well suited for repair of aperforated esophagus with wide drainage of the medias-tinum when diagnosed early. Because most patients with

Accepted for publication Jan 26, 2011.

Address correspondence to Dr Kim, Department of Thoracic and Cardio-vascular Surgery, Pusan National University School of Medicine, Gu-deok-ro Seo-gu, Busan, 602-739, Republic of Korea; e-mail:

this problem are in poor general condition, excessivetrauma should be avoided in these patients to minimizepostoperative complications [8]. The outcomes of treat-ment in patients with Boerhaave’s syndrome were re-viewed in this study to investigate the usefulness ofthoracoscopic surgery.

Patients and Methods

From December 2004 to May 2010, 15 patients withBoerhaave’s syndrome presented to our department.This study includes only patients with Boerhaave’s syn-drome; therefore, if the esophageal perforation wascaused by instrumentation, foreign bodies, externaltrauma, or an underlying disorder such as neuromotordisease or esophageal cancer, these cases were excluded.We have performed thoracoscopic esophageal repair forBoerhaave’s syndrome since 2004. The medical recordswere reviewed retrospectively for preoperative signs andsymptoms, interval between perforation and surgery,surgical time and methods, and outcomes of treatment(hospitalization, postoperative leakage, in-hospital mor-tality, and postoperative complications). The patientswere divided into two groups according to the surgicalapproach (group A, thoracoscopy; group B, thoracotomy).

This study was approved by the institutional review

0003-4975/$36.00doi:10.1016/j.athoracsur.2011.01.082

devia

1553Ann Thorac Surg CHO ET AL2011;91:1552–5 THORACOSCOPIC ESOPHAGEAL REPAIR

GEN

ERA

LT

HO

RA

CIC

board of Pusan National University, Busan, Republic ofKorea. Informed consent was not required for this retro-spective study.

Initial Management and Operative TechniqueInitially, all patients received hydration and broad-spectrum antibiotics and were treated with chest tubedrainage for hydropneumothorax. Regardless of the timeinterval since perforation, emergency surgery was per-formed with primary repair. Although we did not have adefinite indication for thoracoscopic surgery in Boer-haave’s syndrome, when the blood pressure, heart rate,and Sao2 were stable before operation, we consideredsome of these patients as candidates for thoracoscopicsurgery, which was the preferred surgical approach (Ta-ble 1).

In all patients, the inflammatory and necrotic tissue aswell, as the purulent material in the pleural cavity aroundthe esophageal perforation site, was removed and thestatus of the esophageal perforation was determined.Before the repair, the muscular layer was incised toensure that the entire length of the mucosal defect couldbe visualized clearly. Interrupted sutures with absorb-able polyfilament 4-0 thread were provided at the site ofmucosal perforation. The repair was tested by injectingair into a nasogastric tube with occlusion of the distalesophagus while under saline solution. Interrupted su-tures were then provided in the muscular layer usingnonabsorbable polyfilament 3-0 thread. Other para-esophageal and mediastinal spaces were explored, thepleural spaces and mediastinum were copiously irri-gated, and two drain catheters were inserted, with onecatheter placed close to the esophageal suture line alongthe diaphragm for effective drainage.

In group A, four incisions were made to performthoracoscopic surgery: the first incision was made in theseventh intercostal space (ICS) of the midaxillary line toinsert the thoracoscope; the second incision was made inthe sixth ICS of the anterior axillary line; the thirdincision was made in the eighth ICS of the midaxillaryline to insert thoracoscopic instruments, and the fourthincision was made in the ninth ICS of the postaxillary line

Table 1. Preoperative Characteristics of Patients

Group A (mean

Age (years) 52.0 � 8.Interval (hours) 43.5 � 23White blood cell count (103/�L) 10.5 � 8.Hemoglobin (g/dL) 15.1 � 1.Platelet count (103/�L) 185.6 � 20C-reactive protein 16.7 � 16Systolic blood pressure (mm Hg) 111.43 � 10Heart rate (beats/minute) 81.7 � 9.Spo2 (%) 96.7 � 1.Body temperature (°C) 37.1 � 1.

Interval � Interval between perforation and surgery; SD � standard

to insert Endo Retract II (Auto Suture, Tyco Healthcare

Group, Norwalk, CT), for retraction of the diaphragm. Ingroup B, conventional thoracotomy was performed, andbuttress sutures were used in two patients with delayeddiagnoses who had significant necrosis of the esophagus:a diaphragmatic flap in one patient and an omental flapin the other patient to secure the primary suture line andprovide a secondary barrier against possible postopera-tive leakage.

Postoperative ManagementAll patients received nutrition parenterally. Systemicantibiotics were provided until removal of the chest tube.The nasogastric tube was left in place until the seventhpostoperative day; the patient then underwent anesophagography. An oral diet was resumed graduallyonce the integrity of the esophagus was ensured. Inpatients with postoperative leaks noted after esophago-graphy, a second esophagogram was obtained when theamount of chest tube drainage had decreased andcleared.

Regular follow-up was performed in the outpatientclinic after discharge over 1 to 2 months. When relatedsymptoms occurred after the initial follow-up, the pa-tients returned for additional evaluation. The relatedsymptoms, such as dysphagia or gastroesophageal reflux,were assessed by interview during the follow-up periodin the outpatient clinic or by phone calls directly to thepatients until June 2010.

StatisticsData are reported as the mean (range) or as proportions.All data were analyzed with SPSS version 12.0 software(SPSS, Inc, Chicago, IL). Comparisons of the two groupswere performed with the Mann-Whitney and �2 tests forthe variables of interest. The postoperative C-reactiveprotein (CRP) curves of the two groups were analyzed bythe repeated measures of analysis of variance (ANOVA).

Results

All patients were men, with a mean age of 53.1 years(range, 39 to 71 years). All patients underwent primary

D) Group B (mean � SD) p Value

54.1 � 10.6 0.68439.3 � 34.1 0.7868.6 � 5.7 0.620

14.9 � 2.0 0.841196.5 � 78.8 0.715

14.0 � 14.9 0.74595.0 � 16.9 0.046

111.8 � 21.1 0.00493.6 � 6.1 0.19937.2 � 0.9 0.865

tion; Spo2 � oxygen saturation as measured by pulse oximetry.

� S

8.2

84.2.4.77681

esophageal repair; seven patients underwent a thoraco-

0tbgepglwt3h

a

iwgp

ApomamelgwoedTad64

c

T

1554 CHO ET AL Ann Thorac SurgTHORACOSCOPIC ESOPHAGEAL REPAIR 2011;91:1552–5

GEN

ERA

LT

HO

RA

CIC

scopic approach (group A) and eight patients had athoracotomy (group B). Regarding the preoperative char-acteristics of the two groups, systolic blood pressure andheart rate were significantly different and the rest werenot significantly different (Table 1). All patients experi-enced forcible vomiting preceding the onset of symp-toms. The common signs and symptoms identified in-cluded chest or epigastric pain, dyspnea, fever,hematemesis, tachycardia, and hypotension; tachycardiaand hypotension were observed only in group B (Table2). Chest radiographs were obtained in all patients andshowed mediastinal widening, hydropneumothorax, andsubcutaneous emphysema. Chest computed tomographywas carried out in all patients. Endoscopy and esopha-gography were performed to confirm a definite diagnosisin ten patients and one patient, respectively.

The mean interval between perforation and surgerywas 43.5 hours (range, 18.0 to 78.0 hours) in group A and39.3 hours (range, 16.5 to 117.0 hours) in group B (p �.79). Pleural contamination was more severe in group Bhan in group A, but there was no significant differenceetween the two groups, except in several patients inroup B. In particular, there was no severe necrosis of thesophagus that could not be repaired and no visceralleural thickening that needed wide decortication inroup A. The perforation was located exclusively in the

ower third of the esophagus; however, in one patient itas detected from the lower thoracic to the upper part of

he abdominal esophagus. The mean operative time was.7 hours (range, 2.5 to 4.5 hours) in group A and 5.3ours (range, 4.0- to 7.5 hours) in group B (p � 0.005). One

patient in group B died during hospitalization. Postoper-ative ventilator support was needed for one patient ingroup A for 4 hours and in four patients in group B foran average of 106.8 hours (range, 11 to 148 hours). Themean hospitalization was 36.9 days (range, 13 to 73 days)in group A and 38.5 days (range, 18 to 57 days) in groupB (p � 0.73). The postoperative CRP levels for groups A

nd B are shown in Figure 1.Complications occurred in one patient in group A and

n five patients in group B as follows: postoperative leaksere confirmed by esophagography in one patient inroup A and in two patients in group B; dysphagia wasresent in one patient with postoperative leaks in group

able 2. Preoperative Signs and Symptoms

Group A(n � 7)

Group B(n � 8) p Value

Pain (chest, epigastric, orabdominal)

4 8 0.282

Dyspnea 1 3 0.569Hematemesis 2 2 1.000Fever (body temperature �

37.5°C)1 2 1.000

Tachycardia (heart rate �100 beats/minute)

0 6 0.007

Hypotension (systolic blood 0 3 0.200

pressure � 90 mm Hg)

; and pneumonia with empyema occurred in threeatients in group B (Table 3). Three patients with post-perative leaks were treated by conservative manage-ent. Symptoms of the patient with dysphagia were mild

nd he did not need any intervention such as a pneu-atic balloon procedure. Postoperative pneumonia with

mpyema developed in three patients; they had fever,eukocytosis, and pulmonary infiltrates on chest radio-raphs and underwent thoracoscopic pleural irrigation,ith purulent discharge from the chest tube. However,ne patient did not recover from pneumonia with empy-ma and progressed to acute respiratory distress syn-rome (ARDS). The patient died from sepsis and ARDS.here were no other serious complications such as atrialrrhythmia, or other respiratory complications. The me-ian follow-up duration was 40.9 months (range, 0.9 to5.0 months) in group A and 25.6 months (range, 1.0 to6.8 months) in group B (p � 0.25). No patient com-

plained of dysphagia at the last follow-up appointment.

Comment

The management of Boerhaave’s syndrome remains adifficult problem, and early diagnosis is the key to asuccessful outcome. Most surgeons suggest that primaryrepair of a perforated esophagus is the treatment ofchoice in patients with Boerhaave’s syndrome when thecondition is recognized early [6–8]. Furthermore, be-ause the general condition of these patients tends to be

Fig 1. Curve of postoperative C-reactive protein (CRP) levels. (Preop �preoperative day; POD � postoperative day.)

Table 3. Postoperative Complications

Group A (n � 7) Group B (n � 8)

Postoperative leaks 1 2Dysphagia 1 0Postoperative bleeding 0 0

Pneumonia with empyema 0 3

1555Ann Thorac Surg CHO ET AL2011;91:1552–5 THORACOSCOPIC ESOPHAGEAL REPAIR

GEN

ERA

LT

HO

RA

CIC

poor, excessive trauma should be avoided to minimizepostoperative complications. Some investigators recom-mend minimally invasive methods for the treatment ofBoerhaave’s syndrome [8–10]. Tolen and colleagues [9]reported on laparoscopic surgery for Boerhaave’s syn-drome. Landen and El Nakadi [10] suggested a combina-tion of minimally invasive techniques, including laparos-copy, thoracoscopy, mediastinoscopy, and endoscopicstenting. However, most surgeons have been hesitant toperform thoracoscopic surgery to repair a perforatedesophagus, especially in emergency cases.

Herein, the results of thoracoscopic surgery for Boer-haave’s syndrome are presented. There are many bene-fits associated with thoracoscopic repair. First, with tho-racoscopic magnification the proximal and distal edges ofthe longitudinal tear can be visualized more clearly thanwith conventional thoracotomy. As with other types ofminimally invasive surgery, blood loss is minimized, thepostoperative pain is reduced, and the operative trauma(or inflammation) is minimized, whereas ventilation isimproved. Moreover, when a conventional thoracotomyis required, thoracoscopic surgery can be converted toconventional thoracotomy immediately and easily.

However, when preseptic conditions—such as severetachycardia and hypotension with signs of severe infec-tion—are observed before surgery, minimally invasivesurgery might not be suitable, so thoracoscopic surgerywas performed in patients with stable vital signs in ourdata. Although the preoperative condition of patientswith Boerhaave’s syndrome is related to the intervalbetween perforation and surgery, the preoperative con-dition appeared to be more important than the intervalbetween the perforation and surgery for deciding onthoracoscopic repair in our series. Most of the patients ingroup A had relatively mild inflammation of the pleuraand mediastinum. Although there was no patient withsevere necrosis of the esophagus or pleural inflammationwith visceral pleural thickening that needed wide decor-tication in group A, if severe necrosis of the esophagealwall was present in group A, we might have converted toconventional thoracotomy for the buttress procedure anddebridement of necrotic tissue around the perforatedesophagus. The average operative time for group A wassignificantly shorter than that for group B. If inflamma-tion of the pleura and mediastinum was not severe,thoracoscopic surgery was not a time-consuming proce-dure. A postoperative leak was detected in only onepatient in group A, and the leak was treated by conser-vative management without the need for a second oper-ation. Although there was no significant difference in thecurve for postoperative CRP levels between the twogroups (p � 0.97), the increase in the CRP on the firstpostoperative day was relatively small in group A com-

pared with group B, and only one patient required

ventilator care for more than four hours in group A.These findings might indicate that thoracoscopic repaircaused less trauma to patients, which is consistent withother types of minimally invasive surgery. However, theaverage hospitalization of group A patients was notsignificantly shorter than that of group B (average 36.7days versus 38.5 days, p � 0.728).

The results of this study do not definitively confirmthat thoracoscopic esophageal repair is superior to con-ventional esophageal repair in patients with Boerhaave’ssyndrome. There was a selection bias with regard to thepatients included because thoracoscopic repair was usedfor patients in a relatively better preoperative condition.However, the results of this study suggest that primarythoracoscopic esophageal repair is a good surgical alter-native for patients with relatively stable vital signs ormild inflammation in Boerhaave’s syndrome, regardlessof the time interval between the perforation and surgery.A prospective randomized multicenter study is needed toconfirm the efficacy of thoracoscopic esophageal repair inpatients with Boerhaave’s syndrome.

This work was supported for 2 years by Pusan National Univer-sity Research Grant.

References

1. Pate JW, Walker WA, Cole FH, Owen EW, Johnson WH.Spontaneous rupture of the esophagus: a 30-year experi-ence. Ann Thorac Surg 1989;47:689–92.

2. Jones WG, Ginsberg RJ. Esophageal perforation: a continu-ing challenge. Ann Thorac Surg 1992;53:534–43.

3. Lawrence DR, Ohri SK, Moxon RE, Tonsend ER, FountainSW. Primary esophageal repair for Boerhaave’s syndrome.Ann Thorac Surg 1999;67:818–20.

4. Walker WS, Cameron EWJ, Walbaum PR. Diagnosis andmanagement of spontaneous transmural rupture of theoesophagus (Boerhaave’s syndrome). Br J Surg 1985;72:204–7.

5. Pate JW, Walker WA, Cole FH, Owen EW, Johnson WH.Spontaneous rupture of the esophagus: a 30-year experi-ence. Ann Thorac Surg 1989;47:689–92.

6. Jougon J, Bride TM, Delcambre F, Minniti A, Velly JF.Primary esophageal repair for Boerhaave’s syndrome what-ever the free interval between perforation and treatment.Eur J Cardiothorac Surg 2004;25:475–9.

7. Cho S, Jheon S, Ryu KM, Lee EB. Primary esophageal repairin Boerhaave’s syndrome. Dis Esophagus 2008;21:660–3.

8. Ikeda Y, Niimi M, Sasaki Y, Shatari T, Takami H, Kodaira S.Thoracoscopic repair of a spontaneous perforation of theesophagus with the endoscopic suturing device. J ThoracCardiovasc Surg 2001;121:178–9.

9. Tolen C, Hendrickx L, Van Hee R. Laparoscopic treatment ofBoerhaave’s syndrome: a case report and review of theliterature. Acta Chir Belg 2007;107:402–4.

10. Landen S, El Nakadi I. Minimally invasive approach to

Boerhaave’s syndrome: a pilot study of three cases. SurgEndosc 2002;16:1354–7.