J THoRAc CARDIOVASC SURG 85:72-80, 1983 Esophagectomy without thoracotomy: A dangerous operation? Transhiatal esophagectomy without thoracotomy has been performed in 143 patients: 43 with benign disease and 100 with carcinomas at various levels of the esophagus (31 cervicothoracic, five upper third, 33 middle third, and 31 distal third). Esophageal resection and reconstruction were performed in a single stage in 138 patients, and the esophageal substitute was positioned in the posterior mediastinum in the original esophageal bed in 134 patients. Stomach was used to replace the esophagus in 128 patients (93%) and colon in 10 patients. The operative mortality has been 8%, the causes of death being myocardial infarction (three), respiratory insufficiency (three), innominate artery rupture (two), sepsis from mediastinal or retroperitoneal abscess (two), and pulmonary embolus (one). No death was the direct result of the technique of esophagectomy. Complications included intraoperative pneumothorax (51%), transient hoarseness (37%), anastomotic leak (12%), chylothorax (3%), and tracheal laceration (1%). Average intraoperative blood loss for the entire group has been 1,150 ml, 1,800 ml for those requiring concomitant laryngectomy and 900 ml for those undergoing esophagectomy without laryngectomy. Of 63 patients surviving resection of intrathoracic esophageal carcinomas, 86% were discharged, able to swallow, within 3 weeks of operation. Distant lymph node metastases or local tumor invasion precluded a curative resection in 70% of our patients with carcinoma, and the overall average duration of survival has been only 12.5 months. However, of 15 surviving patients with intrathoracic esophageal carcinoma who had "curative" resections, 10 are alive and tumor free from 8 to 60 months (average 31 months) postoperatively. A thoracic incision is seldom required to resect the esophagus for either benign or malignant disease. Transhiatal esophagectomy without thoracotomy is a safe, well-tolerated operation, the "hazards" of which can be minimized by careful technique and experience. Mark B. Orringer, M.D., and Jay S. Orringer, M.D. (by invitation), Ann Arbor, Mich. Our continuing efforts to reduce the morbidity and mortality rates for esophageal resection and recon- struction prompted us to advocate the technique of transhiatal esophagectomy without thoracotomy in pa- tients with both benign and malignant disease requiring esophageal replacement.' By eliminating the need for a thoracotomy, this procedure reduces the operative physiological insult to the patient, and use of a cervical esophageal anastomosis is not associated with the di- sastrous results of disruption of an intrathoracic anas- tomosis. Our preliminary report describing transhiatal esoph- agectomy in 26 patients was criticized for advocating a From the Section of Thoracic Surgery, Department of Surgery, Uni- versity of Michigan Medical Center, Ann Arbor, Mich. Read at the Sixty-second Annual Meeting of The American Associa- tion for Thoracic Surgery, Phoenix, Ariz., May 3-5, 1982. Address for reprints: Mark B. Orringer, M.D., University of Michi- gan Medical Center, Section of Thoracic Surgery, C7079, Box 32, Ann Arbor, Mich. 48109. 72 dangerous operation which violated the basic surgical principles of adequate exposure and hemostasis. This paper presents our cumulative clinical experience with this operation in 143 patients requiring esophagectomy. Our growing facility with this technique and the im- proved operative results in these patients have provided us ample justification for our current belief that trans- hiatal esophagectomy without thoracotomy is the pre- ferred approach in virtually all patients requiring esophageal resection. Patients During the past 5 years, transmediastinal esophagec- tomy without thoracotomy, as described previously,' has been performed in 143 patients, of whom 43 (30%) had benign disease necessitating esophageal replace- ment and 100 (70%) had carcinoma (Table I). The pa- tients with benign disease included 12 men (28%) and 31 women (72%), ranging in age from 18 to 86 years (average 48 years); 28 (65%) had lost from 4.5 to 45 kg (average 10 kg). Seventy-seven (77%) of the patients 0022-5223/83/010072+09$00.90/0 © 1983 The C. V. Mosby Co.
Esophagectomy without thoracotomy: A dangerous operation?
Jan 30, 2023
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Esophagectomy without thoracotomy: A dangerous operation?J THoRAc CARDIOVASC SURG 85:72-80, 1983
Esophagectomy without thoracotomy: A dangerous operation?
Transhiatal esophagectomy without thoracotomy has been performed in 143 patients: 43 with benign disease and 100 with carcinomas at various levels of the esophagus (31 cervicothoracic, five upper third, 33 middle third, and 31 distal third). Esophageal resection and reconstruction were performed in a single stage in 138 patients, and the esophageal substitute was positioned in the posterior mediastinum in the original esophageal bed in 134 patients. Stomach was used to replace the esophagus in 128 patients (93%) and colon in 10 patients. The operative mortality has been 8%, the causes of death being myocardial infarction (three), respiratory insufficiency (three), innominate artery rupture (two), sepsis from mediastinal or retroperitoneal abscess (two), and pulmonary embolus (one). No death was the direct result of the technique of esophagectomy. Complications included intraoperative pneumothorax (51%), transient hoarseness (37%), anastomotic leak (12%), chylothorax (3%), and tracheal laceration (1%). Average intraoperative blood loss for the entire group has been 1,150 ml, 1,800 ml for those requiring concomitant laryngectomy and 900 ml for those undergoing esophagectomy without laryngectomy. Of 63 patients surviving resection of intrathoracic esophageal carcinomas, 86% were discharged, able to swallow, within 3 weeks of operation. Distant lymph node metastases or local tumor invasion precluded a curative resection in 70% of our patients with carcinoma, and the overall average duration of survival has been only 12.5 months. However, of 15 surviving patients with intrathoracic esophageal carcinoma who had "curative" resections, 10 are alive and tumor free from 8 to 60 months (average 31 months) postoperatively. A thoracic incision is seldom required to resect the esophagus for either benign or malignant disease. Transhiatal esophagectomy without thoracotomy is a safe, well-tolerated operation, the "hazards" of which can be minimized by careful technique and experience.
Mark B. Orringer, M.D., and Jay S. Orringer, M.D. (by invitation), Ann Arbor, Mich.
Our continuing efforts to reduce the morbidity and mortality rates for esophageal resection and recon- struction prompted us to advocate the technique of transhiatal esophagectomy without thoracotomy in pa- tients with both benign and malignant disease requiring esophageal replacement.' By eliminating the need for a thoracotomy, this procedure reduces the operative physiological insult to the patient, and use of a cervical esophageal anastomosis is not associated with the di- sastrous results of disruption of an intrathoracic anas- tomosis.
Our preliminary report describing transhiatal esoph- agectomy in 26 patients was criticized for advocating a
From the Section of Thoracic Surgery, Department of Surgery, Uni- versity of Michigan Medical Center, Ann Arbor, Mich.
Read at the Sixty-second Annual Meeting of The American Associa- tion for Thoracic Surgery, Phoenix, Ariz., May 3-5, 1982.
Address for reprints: Mark B. Orringer, M.D., University of Michi- gan Medical Center, Section of Thoracic Surgery, C7079, Box 32, Ann Arbor, Mich. 48109.
72
dangerous operation which violated the basic surgical principles of adequate exposure and hemostasis. This paper presents our cumulative clinical experience with this operation in 143 patients requiring esophagectomy. Our growing facility with this technique and the im- proved operative results in these patients have provided us ample justification for our current belief that trans- hiatal esophagectomy without thoracotomy is the pre- ferred approach in virtually all patients requiring esophageal resection.
Patients
During the past 5 years, transmediastinal esophagec- tomy without thoracotomy, as described previously,' has been performed in 143 patients, of whom 43 (30%) had benign disease necessitating esophageal replace- ment and 100 (70%) had carcinoma (Table I). The pa- tients with benign disease included 12 men (28%) and 31 women (72%), ranging in age from 18 to 86 years (average 48 years); 28 (65%) had lost from 4.5 to 45 kg (average 10 kg). Seventy-seven (77%) of the patients
0022-5223/83/010072+09$00.90/0 © 1983 The C. V. Mosby Co.
IVolume 85 , Number 1
Table I. Indications for transhiatal esophagectomy (143 patients)
underwent immediate esophageal replacement. Partial or total gastric resections were required in four of eight patients with caustic injuries. Colon was used to re- place the esophagus in three patients with caustic in- juries and in patients with carcinoma only when prior gastric resection for peptic ulcer disease made the stomach an unsuitable esophageal substitute. The vis- ceral esophageal substitute was positioned within the posterior mediastinum in the original esophageal bed in all but four patients, in whom either residual posterior mediastinal tumor or fibrosis and narrowing prevented adequate positioning of the stomach for a tension-free cervical anastomosis. In these latter four patients, the retrostemal route was used. In every patient with a nor- mal-sized stomach, the gastric fundus readily reached to the neck for a tension-free cervical anastomosis.
Total gross tumor removal without microscopic evi-
withcarcinoma were men and 23 (23%) women, rang- ing in age from 38 to 92 years (average 61 years); 69 (69%) had lost from 2.3 to 29.5 kg (average 10.5 kg). Sixty-six patients had squamous cell carcinoma of the esophagus. Additional tumors included a thyroid carci- noma involving the cervicothoracic esophagus, an adenosquamouscarcinoma of the upper thoracic esoph- agus, middle-third adenocarcinoma arising in a colum- nar-lined (Barrett's) esophagus (five cases), and distal third adenocarcinoma (27 cases). The 31 patients with cervicothoracic esophageal carcinoma underwent la- ryngopharyngectomy, transhiatal esophagectomy, and either pharyngogastrostomy or pharyngocolostomy; in 20 (65%) an anterior mediastinal tracheostomy, as de- scribed previously," was required because of the need to divide the trachea behind the sternum.
Neuromotor esophageal dysfunction was the com- monest benign condition necessitating esophagectomy, all 20 patients with motor disorders having undergone prior esophageal operations which ultimately failed." The six patients with reflux strictures were not believed to be candidates for intraoperative dilation and Collis- gastroplasty-fundoplication operations, two having chronically perforating Barrett's ulcers, two being men- tally retarded and severely debilitated, and two having undergone multiple antireflux operations. All four pa- tients with acute caustic injuries underwent emergency transhiatal esophagectorny, cervical esophagostomy, and feeding jejunostomy; in three esophageal recon- struction was performed 2 to 8 weeks later. One pa- tient, who had undergone an antiperistaltic retrosternal bypass of the stenosed esophagus following a caustic injury, had recurrent aspiration pneumonia from re- versed retrosternal colonic emptying and spasm. The retrosternal colon was resected, a transmediastinal esophagectomy performed, and a cervical esophago- gastric anastomosis constructed.
Mediastinal inflammation from prior esophageal operations, perforations, or radiation therapy has not been a contraindication to transhiatal esophagectomy. Thirty-six (84%) of our 43 patients with benign disease have had previous procedures, including hiatal hernia repair (18 patients), esophageal dilatation (16 patients), thoracic esophagomyotomy (10 patients), repair of esophageal atresia in infancy (three patients), radiation therapy (two patients), repair of perforation (two pa- tients), and varying reconstructive procedures (six patients).
In all but five patients, esophageal resection and re- construction were performed at the same operation (Table II). The stomach was used as the visceral esophageal substitute in 128 (93%) of our patients who
Benign conditions Neuromotor dysfunction
With gastroesophageal reflux Spasm 4 Achalasia 3 Scleroderma 3 Esophageal atresia 2
Without gastroesophageal reflux Spasm 6 Achalasia 2
Chronic stricture Distal third
Barrett's ulcer-chronic perforation 2 Peptic stricture in mentally retarded 2 Recurrent peptic stricture 2 Following instrumental perforation I
Caustic ingestion Idiopathic upper third stricture Radiation Postemetic stricture Monilial Following laryngopharyngectomy recon-
struction Acute injury
Malfunctioning substernal colon
Total
No.
5 4' I
The Journal of
Thoracic and Cardiovascular
Totals
32 66 98 31 63 94
I 3 4 2 28 30 4 3 7
3 3 3 3 2 2
43 100 143
Type of resection
volvementt vasion:j:
Tumor site No. No. % No. I % No. I %
Intrathoracic Upper third 5 I 20 2 40 3 60 Mid third 33 9 27 9 27 15 45 Distal third 31 7 23 19 61 6 19
Subtotal 69 17 25 30 43 24 38
Cervicothoracic 31 13 42 8 29 12 39
Total 100 30 30 38 38 36 36
*Total gross tumor removal without microscopic vascular or lymphatic inva- sion.
t"Distant" lymph node metastases (e.g .• celiac axis or cervical lymph node involvement by intrathoracic carcinomas or mid-mediastinal nodal metas- tases from cervicothoracic carcinomas).
:j:Gross adherence to prevertebral fascia or tracheobronchial tree, or micro-
scopic vascular or lymphatic invasion.
dence of vascular or lymphatic tumor invasion justified a designation of "curative" resection in only 25% of patients with intrathoracic esophageal carcinomas and in the 42% of patients with cervicothoracic esophageal malignant tumors; thus the overall' 'curative" resection rate was 30% (Table Ill). Postoperative radiation or chemotherapy or both were generally used in patients who had palliative resections.
Results
Complications. There were no intraoperative deaths. One patient with scleroderma, an esophageal stricture, and severe nutritional cirrhosis lost 6,000 rnl
of blood during the operation following a retractor in- jury of the soft, enlarged liver. Measured intraoperative blood loss among the remaining 142 patients averaged 1,150 ml, 1,800 ml for those requiring concomitant laryngectomy and 900 ml for those undergoing esoph- agectomy without laryngectomy. A splenectomy was required because of intraoperative injury in 15 patients (l 0%). Pneumothorax, recognized and treated in- traoperatively with a chest tube, occurred in 73 patients (51%). Left recurrent laryngeal nerve paresis has oc- curred in 37% of patients undergoing a cervical anas- tomosis and has resolved spontaneously within 2 to 12 weeks of operation in all but three patients; these three required a Teflon injection of the vocal cord. Postop- erative chyle fistulas occurred in six patients. Two of the fistulas were cervical and followed laryngopharyn- gectomy; both were controlled with wound catheter suction and a low-residue elemental diet via the jeju- nostomy tube. A thoracotomy was performed in the four patients who had a chylothorax, and the injured thoracic duct was ligated successfully. Iatrogenic hy- poparathyroidism (in eight of 31 laryngopharyngec- tomies) and postoperative innominate artery rupture (in two of 20 anterior mediastinal tracheostomies) have occurred only in our patients undergoing radical cervi- cothoracic esophageal resections.
There were two intraoperative tracheal lacerations. One involved the high membranous trachea and was exposed and repaired through a partial upper sternal split. The other tear involved the membranous carina and was managed by guiding the endotracheal tube into the left main-stem bronchus, ventilating one lung, and performing a substernal gastric bypass. Then, through a right thoracotomy, the esophagectomy was completed and the tracheal tear repaired. The patient had an un- eventful postoperative course.
The relationship between the type of anastomosis and the frequency of anastomotic leaks is shown in Table IV. Pharyngeal anastomotic leaks following
Volume 85 Number 1 January, 1983
Esophagectomy without thoracotomy 7 5
Table IV. Anastomotic leaks after transhiatal esophagectomy (138 anastomoses)
Frequency of leaks
Carcinoma Benign Total
Type of reconstruction No. I % No. I % No. I %
Cervical esophagogastrostomy 5/66 8 3/32 9 8/98 8 Posterior mediastinal route 3/63 5 2/31 6 5194 5 Retrostemal route 2/3 67 III 100 3/4 75
Pharyngogastrostomy 8/28 29 0/2 0 8/30 27 Cervical esophagocolic 0/3 0 0/4 0 0/7 0 Pharyngocolic 1/3 33 1/3 33
Totals 14/100 14 3138 8 17/138 12
laryngopharyngectomy occurred in 29%, but like cer- vical leaks, tended to be well controlled with local drainage. Five percent of patients who had a cervical esophagogastric anastomosis after the stomach was positioned in the posterior mediastinum had anasto- motic disruptions, whereas three of four patients with retrosternal gastric interpositions had anastomotic leaks.
Mortality. There were 11 deaths within 30 days of operation for an overall operative mortality of 8%, five after esophagectomy for carcinoma and three in pa- tients being treated for benign conditions. Causes of death included respiratory insufficiency (three), acute myocardial infarction (three), innominate artery rup- ture (two), sepsis from mediastinal or retroperitoneal abscess (two), and pulmonary embolus (one). Two pa- tients died of respiratory insufficiency following mas- sive aspiration of retained intrathoracic gastric contents during the first postoperative week. The two patients who died of sepsis had carcinoma of the middle third of the esophagus. In one, an extensive tumor was frac- tured away from the prevertebral fascia. Because of the residual posterior mediastinal tumor, the stomach was positioned retrostemally. The patient died after ex- periencing an anastomotic leak followed by progressive respiratory insufficiency and sepsis. At postmortem examination, there was a large posterior mediastinal abscess. The other patient had liver metastases at oper- ation; after palliative esophagectomy, deep throm- bophlebitis of the leg and progressive ascites devel- oped, and the patient ultimately died of sepsis. A large retroperitoneal abscess was found at autopsy.
Six additional patients did not leave the hospital alive following esophagectomy. They died from 6 weeks to 4 months later of fulminant carcinomatosis following re- section of cervicothoracic carcinoma (two), hepato- renal syndrome (one), cerebrovascular accident (one),
Table V. Days of hospitalization following transhiatal esophagectomy (123 patients undergoing one-stage resection and reconstruction and leaving hospital alive)
Hospitalization after operation
No. . (days)
Carcinoma Upper third 5 3 60 17 Middle third 28 18 64 5 18 18 Lower third 30 25 83 3 10 13 Cervicothoracic 23 6 26 5 22 24
Benign conditions 37 23 62 6 16 16
Totals 123 75 61 19 15 17
respiratory insufficiency (one), and renal failure (one). Thus the "hospital" mortality among all 143 patients was 12% (17 patients).
Follow-up. The duration of hospitalization follow- ing transhiatal esophagectomy and immediate recon- struction is shown in Table V. The longest hospitaliza- tions followed resection of cervicothoracic esophageal carcinoma, which often necessitated concomitant medi- astinal tracheostomy. Of 63 patients surviving resection of intrathoracic esophageal carcinomas, 86% (54) were discharged, able to swallow, within 3 weeks of opera- tion and 73% (46) within 7 to 14 days. Of 37 patients surviving resection and reconstruction for benign con- ditions, 78% (29) were discharged within 3 weeks and 62% (23) within 7 to 14 days.
Complete follow-up is available for all patients. Of the 86 patients with carcinoma who survived operation and left the hospital alive, 51 (59%) have died from 6 weeks to 38 months (average 9.3 months) after opera-
7 6 Orringer and Orringer
Table VI. Survival after transhiatal esophagectomy for carcinoma
The Journal of Thoracic and Cardiovascular
Surgery
Survival
12 moormore 18moormore 24 mo or more 36 mo or more No. surviv- Average sur-
I I I No. ITumor site ing operation vival (mo) No. % No. % No. % %
Intrathoracic Upper third 5 9.2 9 32 3 II 2 7 I 4 Middle third 28 9.8 9 32 3 II 2 7 I 4 Distal third 30 16.3 13 42 10 32 9 29 5 16
Subtotal 63 12.9 22 35 13 21 II 17 6 10
Cervicothoracic 23 10.7 7 30 5 22 4 17 I 4
Total 86 12.5 29 34 18 21 15 17 7 8
tion, and all but five of these died of carcinomatosis. Two patients have died of aortogastric fistula formation within 3 months of operation. The remaining 35 pa- tients have been followed up from 1 to 60 months (av- erage 16.5 months). Twelve are alive with metastatic disease from I to 32 months (average 10 months) after operation. Twenty-three patients (27%) are alive and tumor free from 1 to 60 months (average 20 months) after esophagectomy.
Among the 63 surviving patients with carcinoma of the intrathoracic esophagus, 35% have lived 12 months; 17%, 24 months; and only 10%, 36 months or more (Table VI). The longest survival has been in patients with distal third tumors, 42% of whom survived 12 months and 29%, 24 months or more. Of 15 surviving patients with intrathoracic esophageal carcinoma who were believed to have a "curative" resection, three have died of carcinomatosis 6 weeks, 18 months, and 38 months after operation and two are alive but with metastases after 5 months and 34 months. The remain- ing patients are alive and tumor free after 8,9, 10, 18, 21,38,47,48,52, and 60 months, respectively. The latter five had distal third adenocarcinomas.
Of the 40 patients with benign disease who survived esophagectomy and visceral esophageal substitution, six have died from I to 41 months (average 14 months) after operation. The causes of death in these patients have been carcinoma (two), myocardial infarction (one), and suicide (three); in the last three patients, the esophagectomies were originally performed for caustic injuries sustained in their first suicide attempts. The remaining 34 patients have been followed up for 1 to 50 months (average 20 months).
Functional results of visceral esophageal substitution. Benign conditions. Among the 34 living patients and
three who have died after a minimum follow-up of 12
months, 32 (86%) have been able to eat regular, unre- stricted diets and are regarded as having a "good" result of visceral esophageal substitution. Five (14%) have had a "poor" result, i.e., frequent abdominal and chest pain, cramping after eating, and early satiety that prevents a normal meal. Four of these latter patients underwent esophagectomy for recurrent esophageal spasm and, in each case, years of chronic abdominal and chest pain, as well as psychiatric therapy, ante- dated the esophageal resection. The fifth patient un- derwent a colonic interposition following multiple un- successful antireflux operations.
All patients undergoing a cervical esophageal anas- tomosis are instructed to return for outpatient anas- tomotic dilatation if any degree of cervical dysphagia occurs after discharge. Thus 24 of 37 patients (65%) with benign disease have had such outpatient dilata- tions for cervical dysphagia, generally between one and three times during the first 6 postoperative months. Of these, however, only five (14%) have true anastomotic strictures which have necessitated regular bougienage. In three of these latter patients, the stomach was anas- tomosed either to scarred pharynx or cervical esoph- agus following either caustic injury (two) or radiation therapy (one). In one patient an anastomotic stricture developed after healing of a leak; the anastomosis was revised at 18 months and dilatation has not been re- quired for 6 months.
Thirteen patients (35%) have admitted experiencing some regurgitation after operation, generally when bending forward after meals. Only four give a history of nocturnal regurgitation, two following cervical esophagogastrostomy and two after cervical esophago- colostomy. No patient has had any pulmonary compli- cations ofreflux. Fourteen patients (38%) have experi- enced transient postvagotomy diarrhea, generally well controlled with diet and medication (diphenoxylate).
Volume85
Number 1 January. 1983
Two patients have required pyloroplasty because of the late development of gastric outlet obstruction after pyloromyotomy.
Carcinoma. Among the 63 surviving patients with carcinoma of the intrathoracic esophagus, 61 (97%) have been able to swallow a regular diet until the time of their death and represent a "good" result of visceral esophageal substitution. Twenty-six (41%) have had outpatient dilatations for postoperative cervical dys- phagia, but only three had true anastomotic stric- turesnecessitating regular bougienage. Twelve patients (21 %) have experienced postoperative regurgitation, but none has had any pulmonary complications of re- flux. Twenty-six (41%) have had transient postva- gotomy diarrhea.
All 23 patients with cervicothoracic carcinomas could eat regular diets after operation. Two patients have pharyngogastric anastomotic strictures that are di- lated regularly, and five have had postvagotomy diarrhea. Regurgitation is far more common after esophagectomy when the upper esophageal sphincter is resected, and this has been experienced by 17 (74%) of our patients surviving laryngopharyngectomy.
Discussion
Transhiatal esophagectomy has been applicable in virtually every patient requiring an esophageal resec- tion, including those with a megaesophagus of achala- sia and those who have had a previous thoracic esoph- agomyotomy. With more difficult dissections, direct exposure of the esophagus is facilitated by small retrac- tors in the diaphragmatic hiatus. Many patients with esophageal carcinoma metastatic to abdominal or cer- vical lymph nodes are found to have a primary tumor that is still grossly confined to the esophagus, mobile within the mediastinum, and resectable through the transhiatal route. In such patients, esophagectomy and a cervical…
Esophagectomy without thoracotomy: A dangerous operation?
Transhiatal esophagectomy without thoracotomy has been performed in 143 patients: 43 with benign disease and 100 with carcinomas at various levels of the esophagus (31 cervicothoracic, five upper third, 33 middle third, and 31 distal third). Esophageal resection and reconstruction were performed in a single stage in 138 patients, and the esophageal substitute was positioned in the posterior mediastinum in the original esophageal bed in 134 patients. Stomach was used to replace the esophagus in 128 patients (93%) and colon in 10 patients. The operative mortality has been 8%, the causes of death being myocardial infarction (three), respiratory insufficiency (three), innominate artery rupture (two), sepsis from mediastinal or retroperitoneal abscess (two), and pulmonary embolus (one). No death was the direct result of the technique of esophagectomy. Complications included intraoperative pneumothorax (51%), transient hoarseness (37%), anastomotic leak (12%), chylothorax (3%), and tracheal laceration (1%). Average intraoperative blood loss for the entire group has been 1,150 ml, 1,800 ml for those requiring concomitant laryngectomy and 900 ml for those undergoing esophagectomy without laryngectomy. Of 63 patients surviving resection of intrathoracic esophageal carcinomas, 86% were discharged, able to swallow, within 3 weeks of operation. Distant lymph node metastases or local tumor invasion precluded a curative resection in 70% of our patients with carcinoma, and the overall average duration of survival has been only 12.5 months. However, of 15 surviving patients with intrathoracic esophageal carcinoma who had "curative" resections, 10 are alive and tumor free from 8 to 60 months (average 31 months) postoperatively. A thoracic incision is seldom required to resect the esophagus for either benign or malignant disease. Transhiatal esophagectomy without thoracotomy is a safe, well-tolerated operation, the "hazards" of which can be minimized by careful technique and experience.
Mark B. Orringer, M.D., and Jay S. Orringer, M.D. (by invitation), Ann Arbor, Mich.
Our continuing efforts to reduce the morbidity and mortality rates for esophageal resection and recon- struction prompted us to advocate the technique of transhiatal esophagectomy without thoracotomy in pa- tients with both benign and malignant disease requiring esophageal replacement.' By eliminating the need for a thoracotomy, this procedure reduces the operative physiological insult to the patient, and use of a cervical esophageal anastomosis is not associated with the di- sastrous results of disruption of an intrathoracic anas- tomosis.
Our preliminary report describing transhiatal esoph- agectomy in 26 patients was criticized for advocating a
From the Section of Thoracic Surgery, Department of Surgery, Uni- versity of Michigan Medical Center, Ann Arbor, Mich.
Read at the Sixty-second Annual Meeting of The American Associa- tion for Thoracic Surgery, Phoenix, Ariz., May 3-5, 1982.
Address for reprints: Mark B. Orringer, M.D., University of Michi- gan Medical Center, Section of Thoracic Surgery, C7079, Box 32, Ann Arbor, Mich. 48109.
72
dangerous operation which violated the basic surgical principles of adequate exposure and hemostasis. This paper presents our cumulative clinical experience with this operation in 143 patients requiring esophagectomy. Our growing facility with this technique and the im- proved operative results in these patients have provided us ample justification for our current belief that trans- hiatal esophagectomy without thoracotomy is the pre- ferred approach in virtually all patients requiring esophageal resection.
Patients
During the past 5 years, transmediastinal esophagec- tomy without thoracotomy, as described previously,' has been performed in 143 patients, of whom 43 (30%) had benign disease necessitating esophageal replace- ment and 100 (70%) had carcinoma (Table I). The pa- tients with benign disease included 12 men (28%) and 31 women (72%), ranging in age from 18 to 86 years (average 48 years); 28 (65%) had lost from 4.5 to 45 kg (average 10 kg). Seventy-seven (77%) of the patients
0022-5223/83/010072+09$00.90/0 © 1983 The C. V. Mosby Co.
IVolume 85 , Number 1
Table I. Indications for transhiatal esophagectomy (143 patients)
underwent immediate esophageal replacement. Partial or total gastric resections were required in four of eight patients with caustic injuries. Colon was used to re- place the esophagus in three patients with caustic in- juries and in patients with carcinoma only when prior gastric resection for peptic ulcer disease made the stomach an unsuitable esophageal substitute. The vis- ceral esophageal substitute was positioned within the posterior mediastinum in the original esophageal bed in all but four patients, in whom either residual posterior mediastinal tumor or fibrosis and narrowing prevented adequate positioning of the stomach for a tension-free cervical anastomosis. In these latter four patients, the retrostemal route was used. In every patient with a nor- mal-sized stomach, the gastric fundus readily reached to the neck for a tension-free cervical anastomosis.
Total gross tumor removal without microscopic evi-
withcarcinoma were men and 23 (23%) women, rang- ing in age from 38 to 92 years (average 61 years); 69 (69%) had lost from 2.3 to 29.5 kg (average 10.5 kg). Sixty-six patients had squamous cell carcinoma of the esophagus. Additional tumors included a thyroid carci- noma involving the cervicothoracic esophagus, an adenosquamouscarcinoma of the upper thoracic esoph- agus, middle-third adenocarcinoma arising in a colum- nar-lined (Barrett's) esophagus (five cases), and distal third adenocarcinoma (27 cases). The 31 patients with cervicothoracic esophageal carcinoma underwent la- ryngopharyngectomy, transhiatal esophagectomy, and either pharyngogastrostomy or pharyngocolostomy; in 20 (65%) an anterior mediastinal tracheostomy, as de- scribed previously," was required because of the need to divide the trachea behind the sternum.
Neuromotor esophageal dysfunction was the com- monest benign condition necessitating esophagectomy, all 20 patients with motor disorders having undergone prior esophageal operations which ultimately failed." The six patients with reflux strictures were not believed to be candidates for intraoperative dilation and Collis- gastroplasty-fundoplication operations, two having chronically perforating Barrett's ulcers, two being men- tally retarded and severely debilitated, and two having undergone multiple antireflux operations. All four pa- tients with acute caustic injuries underwent emergency transhiatal esophagectorny, cervical esophagostomy, and feeding jejunostomy; in three esophageal recon- struction was performed 2 to 8 weeks later. One pa- tient, who had undergone an antiperistaltic retrosternal bypass of the stenosed esophagus following a caustic injury, had recurrent aspiration pneumonia from re- versed retrosternal colonic emptying and spasm. The retrosternal colon was resected, a transmediastinal esophagectomy performed, and a cervical esophago- gastric anastomosis constructed.
Mediastinal inflammation from prior esophageal operations, perforations, or radiation therapy has not been a contraindication to transhiatal esophagectomy. Thirty-six (84%) of our 43 patients with benign disease have had previous procedures, including hiatal hernia repair (18 patients), esophageal dilatation (16 patients), thoracic esophagomyotomy (10 patients), repair of esophageal atresia in infancy (three patients), radiation therapy (two patients), repair of perforation (two pa- tients), and varying reconstructive procedures (six patients).
In all but five patients, esophageal resection and re- construction were performed at the same operation (Table II). The stomach was used as the visceral esophageal substitute in 128 (93%) of our patients who
Benign conditions Neuromotor dysfunction
With gastroesophageal reflux Spasm 4 Achalasia 3 Scleroderma 3 Esophageal atresia 2
Without gastroesophageal reflux Spasm 6 Achalasia 2
Chronic stricture Distal third
Barrett's ulcer-chronic perforation 2 Peptic stricture in mentally retarded 2 Recurrent peptic stricture 2 Following instrumental perforation I
Caustic ingestion Idiopathic upper third stricture Radiation Postemetic stricture Monilial Following laryngopharyngectomy recon-
struction Acute injury
Malfunctioning substernal colon
Total
No.
5 4' I
The Journal of
Thoracic and Cardiovascular
Totals
32 66 98 31 63 94
I 3 4 2 28 30 4 3 7
3 3 3 3 2 2
43 100 143
Type of resection
volvementt vasion:j:
Tumor site No. No. % No. I % No. I %
Intrathoracic Upper third 5 I 20 2 40 3 60 Mid third 33 9 27 9 27 15 45 Distal third 31 7 23 19 61 6 19
Subtotal 69 17 25 30 43 24 38
Cervicothoracic 31 13 42 8 29 12 39
Total 100 30 30 38 38 36 36
*Total gross tumor removal without microscopic vascular or lymphatic inva- sion.
t"Distant" lymph node metastases (e.g .• celiac axis or cervical lymph node involvement by intrathoracic carcinomas or mid-mediastinal nodal metas- tases from cervicothoracic carcinomas).
:j:Gross adherence to prevertebral fascia or tracheobronchial tree, or micro-
scopic vascular or lymphatic invasion.
dence of vascular or lymphatic tumor invasion justified a designation of "curative" resection in only 25% of patients with intrathoracic esophageal carcinomas and in the 42% of patients with cervicothoracic esophageal malignant tumors; thus the overall' 'curative" resection rate was 30% (Table Ill). Postoperative radiation or chemotherapy or both were generally used in patients who had palliative resections.
Results
Complications. There were no intraoperative deaths. One patient with scleroderma, an esophageal stricture, and severe nutritional cirrhosis lost 6,000 rnl
of blood during the operation following a retractor in- jury of the soft, enlarged liver. Measured intraoperative blood loss among the remaining 142 patients averaged 1,150 ml, 1,800 ml for those requiring concomitant laryngectomy and 900 ml for those undergoing esoph- agectomy without laryngectomy. A splenectomy was required because of intraoperative injury in 15 patients (l 0%). Pneumothorax, recognized and treated in- traoperatively with a chest tube, occurred in 73 patients (51%). Left recurrent laryngeal nerve paresis has oc- curred in 37% of patients undergoing a cervical anas- tomosis and has resolved spontaneously within 2 to 12 weeks of operation in all but three patients; these three required a Teflon injection of the vocal cord. Postop- erative chyle fistulas occurred in six patients. Two of the fistulas were cervical and followed laryngopharyn- gectomy; both were controlled with wound catheter suction and a low-residue elemental diet via the jeju- nostomy tube. A thoracotomy was performed in the four patients who had a chylothorax, and the injured thoracic duct was ligated successfully. Iatrogenic hy- poparathyroidism (in eight of 31 laryngopharyngec- tomies) and postoperative innominate artery rupture (in two of 20 anterior mediastinal tracheostomies) have occurred only in our patients undergoing radical cervi- cothoracic esophageal resections.
There were two intraoperative tracheal lacerations. One involved the high membranous trachea and was exposed and repaired through a partial upper sternal split. The other tear involved the membranous carina and was managed by guiding the endotracheal tube into the left main-stem bronchus, ventilating one lung, and performing a substernal gastric bypass. Then, through a right thoracotomy, the esophagectomy was completed and the tracheal tear repaired. The patient had an un- eventful postoperative course.
The relationship between the type of anastomosis and the frequency of anastomotic leaks is shown in Table IV. Pharyngeal anastomotic leaks following
Volume 85 Number 1 January, 1983
Esophagectomy without thoracotomy 7 5
Table IV. Anastomotic leaks after transhiatal esophagectomy (138 anastomoses)
Frequency of leaks
Carcinoma Benign Total
Type of reconstruction No. I % No. I % No. I %
Cervical esophagogastrostomy 5/66 8 3/32 9 8/98 8 Posterior mediastinal route 3/63 5 2/31 6 5194 5 Retrostemal route 2/3 67 III 100 3/4 75
Pharyngogastrostomy 8/28 29 0/2 0 8/30 27 Cervical esophagocolic 0/3 0 0/4 0 0/7 0 Pharyngocolic 1/3 33 1/3 33
Totals 14/100 14 3138 8 17/138 12
laryngopharyngectomy occurred in 29%, but like cer- vical leaks, tended to be well controlled with local drainage. Five percent of patients who had a cervical esophagogastric anastomosis after the stomach was positioned in the posterior mediastinum had anasto- motic disruptions, whereas three of four patients with retrosternal gastric interpositions had anastomotic leaks.
Mortality. There were 11 deaths within 30 days of operation for an overall operative mortality of 8%, five after esophagectomy for carcinoma and three in pa- tients being treated for benign conditions. Causes of death included respiratory insufficiency (three), acute myocardial infarction (three), innominate artery rup- ture (two), sepsis from mediastinal or retroperitoneal abscess (two), and pulmonary embolus (one). Two pa- tients died of respiratory insufficiency following mas- sive aspiration of retained intrathoracic gastric contents during the first postoperative week. The two patients who died of sepsis had carcinoma of the middle third of the esophagus. In one, an extensive tumor was frac- tured away from the prevertebral fascia. Because of the residual posterior mediastinal tumor, the stomach was positioned retrostemally. The patient died after ex- periencing an anastomotic leak followed by progressive respiratory insufficiency and sepsis. At postmortem examination, there was a large posterior mediastinal abscess. The other patient had liver metastases at oper- ation; after palliative esophagectomy, deep throm- bophlebitis of the leg and progressive ascites devel- oped, and the patient ultimately died of sepsis. A large retroperitoneal abscess was found at autopsy.
Six additional patients did not leave the hospital alive following esophagectomy. They died from 6 weeks to 4 months later of fulminant carcinomatosis following re- section of cervicothoracic carcinoma (two), hepato- renal syndrome (one), cerebrovascular accident (one),
Table V. Days of hospitalization following transhiatal esophagectomy (123 patients undergoing one-stage resection and reconstruction and leaving hospital alive)
Hospitalization after operation
No. . (days)
Carcinoma Upper third 5 3 60 17 Middle third 28 18 64 5 18 18 Lower third 30 25 83 3 10 13 Cervicothoracic 23 6 26 5 22 24
Benign conditions 37 23 62 6 16 16
Totals 123 75 61 19 15 17
respiratory insufficiency (one), and renal failure (one). Thus the "hospital" mortality among all 143 patients was 12% (17 patients).
Follow-up. The duration of hospitalization follow- ing transhiatal esophagectomy and immediate recon- struction is shown in Table V. The longest hospitaliza- tions followed resection of cervicothoracic esophageal carcinoma, which often necessitated concomitant medi- astinal tracheostomy. Of 63 patients surviving resection of intrathoracic esophageal carcinomas, 86% (54) were discharged, able to swallow, within 3 weeks of opera- tion and 73% (46) within 7 to 14 days. Of 37 patients surviving resection and reconstruction for benign con- ditions, 78% (29) were discharged within 3 weeks and 62% (23) within 7 to 14 days.
Complete follow-up is available for all patients. Of the 86 patients with carcinoma who survived operation and left the hospital alive, 51 (59%) have died from 6 weeks to 38 months (average 9.3 months) after opera-
7 6 Orringer and Orringer
Table VI. Survival after transhiatal esophagectomy for carcinoma
The Journal of Thoracic and Cardiovascular
Surgery
Survival
12 moormore 18moormore 24 mo or more 36 mo or more No. surviv- Average sur-
I I I No. ITumor site ing operation vival (mo) No. % No. % No. % %
Intrathoracic Upper third 5 9.2 9 32 3 II 2 7 I 4 Middle third 28 9.8 9 32 3 II 2 7 I 4 Distal third 30 16.3 13 42 10 32 9 29 5 16
Subtotal 63 12.9 22 35 13 21 II 17 6 10
Cervicothoracic 23 10.7 7 30 5 22 4 17 I 4
Total 86 12.5 29 34 18 21 15 17 7 8
tion, and all but five of these died of carcinomatosis. Two patients have died of aortogastric fistula formation within 3 months of operation. The remaining 35 pa- tients have been followed up from 1 to 60 months (av- erage 16.5 months). Twelve are alive with metastatic disease from I to 32 months (average 10 months) after operation. Twenty-three patients (27%) are alive and tumor free from 1 to 60 months (average 20 months) after esophagectomy.
Among the 63 surviving patients with carcinoma of the intrathoracic esophagus, 35% have lived 12 months; 17%, 24 months; and only 10%, 36 months or more (Table VI). The longest survival has been in patients with distal third tumors, 42% of whom survived 12 months and 29%, 24 months or more. Of 15 surviving patients with intrathoracic esophageal carcinoma who were believed to have a "curative" resection, three have died of carcinomatosis 6 weeks, 18 months, and 38 months after operation and two are alive but with metastases after 5 months and 34 months. The remain- ing patients are alive and tumor free after 8,9, 10, 18, 21,38,47,48,52, and 60 months, respectively. The latter five had distal third adenocarcinomas.
Of the 40 patients with benign disease who survived esophagectomy and visceral esophageal substitution, six have died from I to 41 months (average 14 months) after operation. The causes of death in these patients have been carcinoma (two), myocardial infarction (one), and suicide (three); in the last three patients, the esophagectomies were originally performed for caustic injuries sustained in their first suicide attempts. The remaining 34 patients have been followed up for 1 to 50 months (average 20 months).
Functional results of visceral esophageal substitution. Benign conditions. Among the 34 living patients and
three who have died after a minimum follow-up of 12
months, 32 (86%) have been able to eat regular, unre- stricted diets and are regarded as having a "good" result of visceral esophageal substitution. Five (14%) have had a "poor" result, i.e., frequent abdominal and chest pain, cramping after eating, and early satiety that prevents a normal meal. Four of these latter patients underwent esophagectomy for recurrent esophageal spasm and, in each case, years of chronic abdominal and chest pain, as well as psychiatric therapy, ante- dated the esophageal resection. The fifth patient un- derwent a colonic interposition following multiple un- successful antireflux operations.
All patients undergoing a cervical esophageal anas- tomosis are instructed to return for outpatient anas- tomotic dilatation if any degree of cervical dysphagia occurs after discharge. Thus 24 of 37 patients (65%) with benign disease have had such outpatient dilata- tions for cervical dysphagia, generally between one and three times during the first 6 postoperative months. Of these, however, only five (14%) have true anastomotic strictures which have necessitated regular bougienage. In three of these latter patients, the stomach was anas- tomosed either to scarred pharynx or cervical esoph- agus following either caustic injury (two) or radiation therapy (one). In one patient an anastomotic stricture developed after healing of a leak; the anastomosis was revised at 18 months and dilatation has not been re- quired for 6 months.
Thirteen patients (35%) have admitted experiencing some regurgitation after operation, generally when bending forward after meals. Only four give a history of nocturnal regurgitation, two following cervical esophagogastrostomy and two after cervical esophago- colostomy. No patient has had any pulmonary compli- cations ofreflux. Fourteen patients (38%) have experi- enced transient postvagotomy diarrhea, generally well controlled with diet and medication (diphenoxylate).
Volume85
Number 1 January. 1983
Two patients have required pyloroplasty because of the late development of gastric outlet obstruction after pyloromyotomy.
Carcinoma. Among the 63 surviving patients with carcinoma of the intrathoracic esophagus, 61 (97%) have been able to swallow a regular diet until the time of their death and represent a "good" result of visceral esophageal substitution. Twenty-six (41%) have had outpatient dilatations for postoperative cervical dys- phagia, but only three had true anastomotic stric- turesnecessitating regular bougienage. Twelve patients (21 %) have experienced postoperative regurgitation, but none has had any pulmonary complications of re- flux. Twenty-six (41%) have had transient postva- gotomy diarrhea.
All 23 patients with cervicothoracic carcinomas could eat regular diets after operation. Two patients have pharyngogastric anastomotic strictures that are di- lated regularly, and five have had postvagotomy diarrhea. Regurgitation is far more common after esophagectomy when the upper esophageal sphincter is resected, and this has been experienced by 17 (74%) of our patients surviving laryngopharyngectomy.
Discussion
Transhiatal esophagectomy has been applicable in virtually every patient requiring an esophageal resec- tion, including those with a megaesophagus of achala- sia and those who have had a previous thoracic esoph- agomyotomy. With more difficult dissections, direct exposure of the esophagus is facilitated by small retrac- tors in the diaphragmatic hiatus. Many patients with esophageal carcinoma metastatic to abdominal or cer- vical lymph nodes are found to have a primary tumor that is still grossly confined to the esophagus, mobile within the mediastinum, and resectable through the transhiatal route. In such patients, esophagectomy and a cervical…
Related Documents
