Ann Thorac Surg 2003;76:903-908
© 2003 The Society of Thoracic Surgeons
Original article: general thoracic
Analysis of pulmonary complications after three-field lymph node dissection for esophageal cancer
Wentao Fang, MDa,
Hoichi Kato, MDa*,
Yuji Tachimori, MDa,
Hiroyasu Igaki, MDa,
Hiroshi Sato, MDa,
Hiroyuki Daiko, MDa
a Department of Surgery, National Cancer Center Hospital, Tokyo, Japan
Accepted for publication March 13, 2003.
* Address reprint requests to Dr Kato, Department of Surgery, National Cancer Center Hospital, 1-1 Tsukiji 5-chome, Chuo-ku, Tokyo 104-0045, Japan
e-mail: hckato{at}ncc.go.jp
 |
Abstract
|
|---|
BACKGROUND: Pulmonary complications are a major component of morbidity and mortality after esophagectomy, and have not been well studied after extended lymphadenectomy.
METHODS: Four hundred forty-one patients underwent three-field lymph node dissection and were retrospectively reviewed. Pulmonary complications developed in 32 patients (7.3%) and resulted in 11 deaths (34.4% of pulmonary complications were fatal, and 62.4% of all mortality was caused by pulmonary complications). Pulmonary complications were divided into primary (group A) and secondary pulmonary morbidities (group B), and analyzed separately. Perioperative arterial blood gases on room air were compared with a matched control group (group C).
RESULTS: All primary complications occurred in the first postoperative week, whereas secondary complications were distributed evenly after operation. The incidence of serious infection (60% versus 23.5%, p = 0.041) and respiratory failure (70.6% versus 31.6%, p = 0.045) was significantly higher in group B as compared with group A and was associated with a higher death rate (47.1% versus 15.8%, p = 0.047). Changes in arterial blood gases were similar in groups A and C, both PaO2 and pH were reduced in group B, and PaCO2 was increased. Independent risk factors for primary pulmonary complications were history of major operation, abnormal spirometry, and chronic renal dysfunction. Predictive factors for secondary pulmonary complications were old age, concomitant total gastrectomy, major anastomotic leakage, and bilateral vocal cord palsy.
CONCLUSIONS: Pulmonary complications can be kept at a low level, but they still account for most of the mortality after three-field lymph node dissection. Primary and secondary pulmonary complications are two distinct entities that should be managed differently. Arterial blood gases on room air are helpful in the management of pulmonary complications.
|
Introduction
|
|---|
Although complication rates after surgery for esophageal cancer have been significantly reduced in recent years [1], pulmonary complications are still considered a major component of morbidity and one of the dominant causes of operative mortality [2–5]. As esophagectomy with extended lymph node dissection is believed to increase the survival of esophageal cancer patients, fear of the increased morbidity rate associated with it has been one of the obstacles preventing its wider application [6, 7]. Among the literature dealing with pulmonary complications after esophagectomy, few have focused on its occurrence after three-field lymph node dissection. The objectives of this study were to identify the risk factors for pulmonary complications to prevent its occurrence, and to improve management of complications when they occur, so that patients with esophageal cancer may benefit more from three-field lymph node dissection.
|
Patients and methods
|
|---|
Records of 441 patients who underwent esophagectomy and three-field lymph node dissection for esophageal cancer at National Cancer Center Hospital, Tokyo, from January 1986 to December 1998 were reviewed. More than 90% of these patients had squamous cell carcinoma located in the thoracic esophagus. A cervicothoracoabdominal approach was used, as described previously [7]. The upper digestive tract was reconstructed using a stomach tube through a retrosternal route with a neck anastomosis. Total gastrectomy and colon reconstruction was performed simultaneously in patients with concomitant gastric cancer. All but 1 patient was extubated immediately after the operation. Flexible fiber bronchoscopy for bronchial lavage and to check vocal cord movement was performed at least daily for a couple of days postoperatively. Arterial blood gases were routinely checked every morning during the first week under room air conditions to avoid the influence of inhaled oxygen concentration. Oral intake was begun on the seventh postoperative day after barium esophagogram had ruled out anastomotic leakage.
Pulmonary complications were defined as follows:
- Pneumonia: fever, productive cough, increased white blood cell count, and marked infiltration on chest roentgenogram, which required prolonged coverage with antibiotics.
- Respiratory failure: deterioration of respiratory status necessitating mechanical ventilatory support.
- Empyema: purulent exudation in the pleural cavity with fever and increased white blood cell count requiring drainage and antibiotic therapy.
- Air leak: prolonged (more than 7 days) air leak from chest tubes refractory to negative pressure drainage necessitating intervention.
Complications occurred in 275 patients (62.4%) after operation, leading to 17 deaths for a total in-hospital mortality of 3.9%. Among them, 32 patients experienced pulmonary complications (7.3% of all 441 patients and 11.6% of total morbidity), and 11 died during hospital stay (34.4% of pulmonary complications and 62.4% of total mortality). The pulmonary complications were divided into two groups, as we considered them different entities that should be analyzed separately: group A, 19 cases having primary pulmonary complications after an uneventful operation without any obvious predisposing factor; and group B, 17 cases having pulmonary complications secondary to other surgical morbidities (including 4 patients in group A who had both primary and secondary complications).
The demographic factors, management, and tumor characteristics of the two groups were compared with those of 409 patients who did not have pulmonary complications (Table 1).
Factors of statistically significant difference were entered as dependent variables into multivariate risk factor analysis. A control group of 32 patients (group C) was selected from the 409 patients without pulmonary complications by matching the patients’ age, tumor stage, and operative procedures with groups A and B so as to compare the results of arterial blood gases among these three groups.
Continuous data were evaluated by Student’s t test, and contingency tables were analyzed using 
2 test or Fisher’s exact test when appropriate. Multivariate analysis of risk factors for pulmonary morbidity was performed with logistic regression. All probabilities were two-tailed, with p values less than 0.05 regarded as statistically significant.
|
Results
|
|---|
Pneumonia and respiratory failure were the two most common pulmonary complications after three-field lymph node dissection, occurring in 28 (6.4%) and 18 (4.1%) patients respectively. Pneumonia occurred in 17 of 19 patients in group A, and in 4 cases (of 17, 23.5%) it was a serious infection caused by antibiotic-resistant bacteria such as methicillin-resistant Staphylococcus aureus or Pseudomonas aeruginosa. Five patients further progressed to respiratory insufficiency, and 3 had empyema. One patient was unable to be weaned from the ventilator after the operation, and another patient with a prolonged air leak had to undergo reexploration. All the complications occurred within the first postoperative week, with its peak incidence in the first 3 days. Altogether, there were a total of 6 cases (31.6%) of respiratory failure in group A, and 3 (15.8%) of them died consequently.
The postoperative course of group B patients was more variable. Eighteen events of pneumonia occurred in 15 patients, among which 8 events were caused by methicillin-resistant S aureus or P aeruginosa infection, and 1 case was septic shock of unknown etiology (9 of 15, 60.0%, p = 0.041 versus group A). Pneumonia developed after systemic infection in 8 patients, because of aspiration in another 8, because of a tracheal fistula in 1 patient, and after reoperation for chylothorax in 1 patient. Pneumonia resulted in respiratory insufficiency in 10 cases, of which 3 were further complicated by empyema. There were another 2 patients in group B who also developed respiratory failure, 1 from hypovolemic shock after massive bleeding during surgery, and the other from leak-induced peritonitis. These complications were distributed quite evenly throughout the first month after operation. There were a total of 12 cases (70.6%, p = 0.045 versus group A) of respiratory failure in group B, with 8 deaths (47.1%, p = 0.047 versus group A) related to pulmonary complications.
Figures 1, 2, and 3
compare the perioperative changes
in arterial blood gases on room air in the three groups. In group C,
PaO2 decreased sharply on the first to third postoperative days, and then gradually returned to a normal level of more than 75 mm Hg. Values for PaCO2 and pH were in the normal range. Group A displayed similar patterns of change as group C; however, in group A, PaO2 dropped significantly lower, recovered more slowly, and failed to return to normal at the end of the first week compared with group C. There was no significant difference in PaCO2 and pH between groups A and C. In contrast, there was no significant difference between PaO2 of groups B and C during the first week, except that PaO2 of group B did not recover as quickly as group C did, so that a significant difference between the two groups was reached by the end of the first week. In addition, PaCO2 of group B was significantly higher than that of group C in the first 3 days, reaching more than 45 mm Hg on the third and forth postoperative days, and correspondingly the pH values decreased significantly.
View larger version (15K):
[in this window]
[in a new window]
|
Fig 1. Comparison of PaO2 changes on room air in the first postoperative week between groups A, B, and C. (Top) Group A (patients with primary pulmonary complications) versus group C (control group). (Bottom) Group B (patients with secondary pulmonary complications) versus group C. (* = statistically significant difference compared with group C.)
|
|
View larger version (15K):
[in this window]
[in a new window]
|
Fig 2. Comparison of PaCO2 changes on room air in the first postoperative week between groups A, B, and C. (Top) Group A (patients with primary pulmonary complications) versus group C (control group). (Bottom) Group B (patients with secondary pulmonary complications) versus group C. (* = statistically significant difference compared with group C.)
|
|
View larger version (15K):
[in this window]
[in a new window]
|
Fig 3. Comparison of pH changes on room air in the first postoperative week between groups A, B, and C. (Top) Group A (patients with primary pulmonary complications) versus group C (control group). (Bottom) Group B (patients with secondary pulmonary complications) versus group C. (* = statistically significant difference compared with group C.)
|
|
Possible risk factors for primary pulmonary complications identified in univariate analysis were old age, history of major operation, chronic pulmonary diseases, chronic renal dysfunction, and abnormal spirometry. When these factors were entered into multivariate analysis, only history of major operation, chronic renal dysfunction, and abnormal spirometry reached statistical significance and were thus regarded as independent risk factors. On the other hand, although old age, history of major operation, chronic renal dysfunction, abnormal spirometry, concomitant total gastrectomy, colon reconstruction, bilateral vocal cord palsy, and major anastomotic leakage were identified as possible risk factors for secondary pulmonary complications by univariate analysis, only old age, concomitant gastrectomy, bilateral vocal cord palsy, and major leakage had independent predictive value on multivariate analysis (Table 2).
|
Comment
|
|---|
The occurrence of pulmonary complications after transthoracic esophagectomy for esophageal cancer has been reported to be 20% to 50%, with a 10% to 40% mortality rate [2–5, 8–12]. Pulmonary complications account for almost half of the morbidity and mortality, and are therefore regarded as the most common cause of operative death [2–5]. The discrepancy between the incidence reported in the literature has been attributed to different definition criteria and perioperative management strategies [12]. According to our definition, the rate of pulmonary complications after esophagectomy with three-field lymph node dissection was only 7.3%, which constituted merely 11.6% of the total morbidity. Although extended lymphadenectomy was assumed to increase the risk of pulmonary problems, our results compare favorably with those previously reported in the literature. Nevertheless, this complication still resulted in death in one third of our patients having pulmonary complications and was related to 62% of total mortality. Therefore, a better understanding of the clinical course and possible risk factors will help reduce the occurrence of pulmonary complications and improve their management when they do occur. In our experience, pulmonary complications in some patients were primary, occurring directly after an uneventful operation, without any obvious predisposing factor, whereas in other patients pulmonary complications were clearly the consequence of a complicated procedure or other postoperative morbidities quite irrelevant to the patients’ respiratory reserve. These two types of complications tend to occur at different times, have different clinical presentations, and carry different outcomes. For this reason, they were regarded as two different entities and analyzed separately in the current study.
The usual change of arterial blood gases in patients without any pulmonary complication was a decrease in PaO2 without any significant change in PaCO2 or acid-base balance, on the first to third postoperative days as a result of relative volume overloading because of refilling of water from the interstitial space as a reaction to the trauma of the surgery. This coincided well with the timing of primary pulmonary complications in our patients, which all happened in the first postoperative week and had their peak incidence in the first 3 days. Most patients in group A presented with symptoms similar to those of acute pulmonary edema. The pathophysiologic feature of this group was a further depression and slower recovery of PaO2 without considerable changes in PaCO2 or acid-base balance. Consequently, the prognosis of this group was less ominous. Respiratory insufficiency occurred in 30% of the patients, and the death rate was 15%.
History of major operation, abnormal spirometry, and chronic renal dysfunction were identified as independent risk factors for primary pulmonary complications in this study. With these three factors, 96.6% of the primary pulmonary complications could be predicted. The first two are in accordance with the previous literature [3–5, 8, 9]. They represent the patients’ general fitness and respiratory reserve. Patients with renal deficiency were less capable of coping with the sudden increase of circulating volume during the refilling period after three-field lymph node dissection and thus were prone to having respiratory troubles [9]. On account of this, precaution should be taken while considering three-field lymph node dissection in patients with a history of operation or chronic renal dysfunction, or in those with abnormal spirometry showing both restrictive and obstructive impairment of pulmonary function. Fluid restriction and application of diuretics are recommended in the first 3 postoperative days, and dopamine may help solve the dilemma of avoiding volume overload while maintaining blood pressure and renal blood flow.
Secondary pulmonary complications in group B were more desperate conditions. These complications could happen at any time during the postoperative course. More than half of the pneumonias in this group were the result of systemic bacterial infection, mostly antibiotic-resistant ones, which were significantly more frequent than those in group A. Another major component was aspiration pneumonia, which often led to respiratory insufficiency or even acute cardiac arrest. The changes of arterial blood gases in group B included both PaO2 depression and PaCO2 elevation along with decreased pH, representing decompensation of the respiratory system from surgical accidents or other serious operative morbidities. In fact, it was this subgroup of complications that had the worst prognosis and accounted for most of the mortality in the current study. More than 70% of the patients progressed to respiratory failure, and almost 50% ultimately died, both being significantly higher than comparable events in group A.
Concomitant gastrectomy, major leakage, bilateral vocal cord palsy, and old age were identified as independent risk factors for secondary pulmonary complications. Concomitant total gastrectomy with colon reconstruction might be the most extensive and time-consuming procedure, carrying the highest morbidity rate [13, 14]. In our study, pulmonary complication rates were similar between patients with or without anastomotic leakage. However, major leaks that failed to respond to conservative measures should be actively managed. Otherwise, patients’ nutritional and immune states are severely affected, causing rapid deterioration of their general condition, and predisposing to serious infection. By routine check of vocal cord movement and early local application of collagen through bronchoscope, we managed to keep the occurrence of pneumonia in patients with unilateral recurrent nerve paralysis at a similar level to those without the problem. However, with bilateral paralysis, aspiration was almost unavoidable. In our patients, bilateral vocal cord palsy mainly originated from resection of the recurrent laryngeal nerve invaded by metastatic lymph nodes, which was detected during operation. There were significantly more aspiration events in these patients, and once aspiration occurred, half of these patients developed respiratory failure. In the current study, the only risk factor for secondary pulmonary complications that could be established before operation was the age of the patients, as elderly individuals tended to have more perioperative complications and were less able to tolerate the surgical stress [15]. We have demonstrated previously that esophagectomy with three-field lymph node dissection can be performed safely in patients older than 70 years of age, with satisfactory long-term result comparable to those in younger patients. However, elderly patients with multiorgan dysfunction have a much higher risk of developing medically related complications such as multiorgan failure and infection [16]. On the basis of this concern, extensive procedures such as concomitant total gastrectomy are not recommended in elderly patients, especially those with multiorgan dysfunction. Extended lymphadenectomy should not be denied solely because of age, but conditions such as bilateral vocal cord palsy should be definitely avoided in elderly patients even if the extent of lymph node dissection is compromised to some extent.
Inasmuch as secondary pulmonary complications carry such a poor prognosis and are difficult to anticipate, early detection and timely intervention of predisposing problems is of critical importance. Apart from prevention of its occurrence, anastomotic leakage refractory to conservative therapy should be managed actively before the patient’s general condition and immune status deteriorate. Routine check of vocal cord movement after operation and, in the presence of bilateral paralysis, preventive minitracheostomy are helpful in reducing the frequency and severity of aspiration [17].
In our experience, arterial blood gases on room air are helpful in directing perioperative management and differentiating primary from secondary pulmonary complications in the first postoperative week. With the knowledge of the normal pattern of change, it is easier to assess the circulation status and to modulate fluid administration during the refilling period so as to minimize its impact on patients’ respiratory condition and, therefore, to prevent primary pulmonary complications. In cases of continued PaO2 depression along with abnormal rise of PaCO2 or even subsequent pH change, serious respiratory problems secondary to other surgical morbidities should be suspected, and careful investigation of possible underlying causes and aggressive intervention should be started promptly before the patient’s condition further deteriorates.
In conclusion, pulmonary complications can be kept at an acceptably low level, but still account for most of the mortality after esophagectomy with three-field lymph node dissection. Primary and secondary pulmonary complications are two distinct entities with different etiology, pathophysiology, clinical presentation, and prognosis. Arterial blood gases on room air are helpful in routine postoperative management and in differentiating respiratory complications in the early postoperative period. Prevention of possible risk factors and aggressive intervention of predisposing problems are important in improving the outcome of pulmonary complications so that patients with esophageal cancer may benefit more from the radical procedure of extended lymph node dissection.
|
Acknowledgments
|
|---|
The authors thank Robert Shen, MD, for revising the English syntax.
|
References
|
|---|
- Muller J.M., Erasmi H., Stelzner M., Zieren U., Pichlmaier H. Surgical therapy of esophageal carcinoma. Br J Surg 1990;77:845-857.[Medline]
- Ferguson M.K., Martin T.R., Reeder L.B., Olla J. Mortality after esophagectomy: risk factor analysis. World J Surg 1997;21:599-604.[Medline]
- Dumont P., Wihlm J.M., Hentz J.G., Roeslin N., Lion R., Morand G. Respiratory complications after surgical treatment of esophageal cancer. A study of 309 patients according to the type of resection. Eur J Cardiothorac Surg 1995;9:539-543.[Abstract]
- Law S.Y.K., Fok M., Wong J. Risk analysis in resection of squamous cell carcinoma of the esophagus. World J Surg 1994;18:339-346.[Medline]
- Liedman B.L., Bennegard K., Olbe L.C., Lundell L.R. Predictors of postoperative morbidity and mortality after surgery for gastro-esophageal carcinomas. Eur J Surg 1995;161:173-180.[Medline]
- Akiyama H., Tsurumaru M., Udagawa H., Kajiyama Y. Radical lymph node dissection for cancer of the thoracic esophagus. Ann Surg 1994;220:364-373.[Medline]
- Kato H., Watanabe H., Tachimori Y., Iizuka T. Evaluation of neck lymph node dissection for thoracic esophageal carcinoma. Ann Thorac Surg 1991;51:931-935.[Abstract]
- Nagawa H., Kobori O., Muto T. Prediction of pulmonary complications after transthoracic esophagectomy. Br J Surg 1994;81:860-862.[Medline]
- Nishi M., Hiramatsu Y., Hioki K., Hatano T., Yamamoto M. Pulmonary complications after subtotal esophagectomy. Br J Surg 1988;75:527-530.[Medline]
- Goldminc M., Maddern G., Le Prise E., Meunier B., Campion P., Launois B. Esophagectomy by a transhiatal approach or thoracotomy: a prospective randomized trial. Br J Surg 1993;80:367-370.[Medline]
- Gluch L., Smith R.C., Bambach C.P., Brown A.R. Comparison of outcomes following transhiatal or Ivor Lewis esophagectomy for esophageal carcinoma. World J Surg 1999;23:271-276.[Medline]
- Crozier T.A., Sydow M., Siewert J.R., Braun U. Postoperative pulmonary complication rate and long-term changes in respiratory function following esophagectomy with esophagogastrostomy. Acta Anaesthesiol Scand 1992;36:10-15.[Medline]
- Cerfolio R.J., Allen M.S., Deschamps C., Trastek V.F., Pairolero P.C. Esophageal replacement by colon interposition. Ann Thorac Surg 1995;59:1382-1384.[Abstract/Free Full Text]
- Zhang D.W., Cheng G.Y., Huang G.J., et al. Operable squamous esophageal cancer: current results from the East. World J Surg 1994;18:347-354.[Medline]
- Poon R.T., Law S.Y.k, Chu K.M., Branicki F.J., Wong J. Esophagectomy for carcinoma of the esophagus in the elderly. Results of current surgical management. Ann Surg 1998;227:357-364.[Medline]
- Fang W., Igaki H., Tachimori Y., Sato H., Daiko H., Kato H. Three-field lymph node dissection for esophageal cancer in elderly patients over 70 years of age. Ann Thorac Surg 2001;72:867-871.[Abstract/Free Full Text]
- Wain J.C., Wilson D.J., Mathisen D.J. Clinical experience with minitracheostomy. Ann Thorac Surg 1990;49:881-886.[Abstract]
This article has been cited by other articles:
|
|
|
|
 
H.-S. Lee, J. M. Lee, M. S. Kim, H. Y. Kim, B. Hwangbo, and J. I. Zo
Low-Dose Steroid Therapy at an Early Phase of Postoperative Acute Respiratory Distress Syndrome
Ann. Thorac. Surg.,
February 1, 2005;
79(2):
405 - 410.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
