| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2003;76:363-369
© 2003 The Society of Thoracic Surgeons
a Department of Thoracic Surgery, Otto Wagner Hospital, Vienna, Austria
Accepted for publication February 12, 2003.
* Address reprint requests to Dr Peter Hollaus, Department of Thoracic Surgery, Otto Wagner Hospital, Sanatoriumstraße 2, A-1140 Vienna, Austria.
e-mail: peter.hollaus{at}pul.magwien.gv.at
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and methodsResultsCommentReferences |
|---|
METHODS: Sex, age, tnm-stage, histology, neoadjuvant chemotherapy, disease-free interval (months), exact localisation of tumour recurrence, time between first and second recurrence (months), survival after first and second recurrence (months), causes of death were retrospectively recorded in 83 patients operated between December 1993 and July 2001.
RESULTS: One patient was lost to follow-up, five resections were nonradical. Survivors’ follow-up lasted 5 to 100.7 months (mean 43.3). Fourteen patients (14.4%) died tumor free. Eleven (13.2%) distant recurrences were diagnosed 1 to 42 months (mean 10.6) postoperatively, eight (9.6%) died 0 to 17 months (mean 7.55) after diagnosis. Nine local recurrences (10.8%)–5 unifocal, 4 multifocal–occurred 2 to 35 months (mean 17.3) postoperatively, eight died 0 to 8 months (mean 2.13) after diagnosis. Nine mixed recurrences (10.8%)–1 synchronous, 8 metachronous–were found (14.8%) 2 to 21 months postoperatively (mean 8.3). All died 4 to 41 months (mean 17.83) after diagnosis. Fourteen mediastinal lymph node recurrences occurred, ten as a primary recurrence and four as secondary. Lymph nodes were involved in all multifocal recurrences. Intrabronchial recurrence was observed in five patients and was always a result of progressive regional lymph node recurrence.
CONCLUSIONS: The pattern and natural history of recurrence cannot be sufficiently explained by stage and surgical radicality and suggest different genetic characteristics of the primary tumor. In case of reoperation due to intrabronchial recurrence adjuvant mediastinal irradiation should be considered.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentReferences |
|---|
|
Patients and methods |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentReferences |
|---|
Postoperative routine follow-up consisted of a bronchoscopy, chest roentgenography, chest computed tomography (CT) every 6 months during the first 2 postoperative years and annual restagings from the third to the fifth postoperative year. Radiologic evidence of recurrence was considered sufficient in case of mediastinal lymph node enlargement (> 1 cm diameter) suggesting mediastinal lymph node recurrence and multiple intra- and extrathoracic lesions suggesting distant metastasis. Histologic specimens were obtained of intrabronchial recurrence, of a single intrapulmonary lesion to exclude secondary carcinoma and of single extrapulmonary lesions if operation was considered.
A recurrence was defined as unifocal if only one anatomic region was involved, as multifocal if two or more different anatomic regions were involved. Local recurrence was defined as any recurrence within the ipsilateral hemithorax. Synchronous and metachronous tumours were classified according to the criteria of Martini and Melamed [5].
The patients’ fate was retrospectively extracted from their hospital charts and ambulatory charts and evaluated by direct contact with the patients, their relatives or their family doctors. The following data were recorded: sex, age, side, TNM stage, histology, neoadjuvant chemotherapy, disease-free interval (months), the exact localisation of tumour recurrence, time between first and second recurrence (months), survival after operation, first and second recurrence (months), causes of death, secondary thoracic interventions.
Statistical analysis
Statistical analysis was performed with SPSS for windows 10.0. (Chicago, IL). All summary statistics are presented as mean ± standard deviation for continuous variables and as percentages for categoric variables. Crosstabulation and Chi-square test were used for comparison of categorical variables. Results were considered to be significant at a p value less than 0.05. To identify significant independent predictors of recurrence, all factors with a significance < 0.05 were entered into multivariate analysis. Odds ratio (OR) and their 95% confidence intervals were calculated using a binary backwards stepwise logistic regression estimates. The combination of independent variables giving the best explanation of the outcome (using R2- statistic) was adopted.
|
Results |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentReferences |
|---|
Technical details
We performed 59 right-sided (71.1%) and 24 left-sided (28.9%) procedures. The tumor was located in the upper, lower, and middle lobe in 63 patients (75.9%), 11 patients (13.3%) and 1 patient (1.2%), respectively. In 5 patients (6%) the upper and the middle lobe, in 3 patients (3.6%) the lower and middle lobe were involved. These patients required bilobectomies. There was no sleeve pneumonectomy.
In 16 patients (19.3%) a wedge resection of the main bronchus was performed, in 55 patients (66.3%) the bronchial tree was reconstructed with an end-to-end anastomsis, and in 12 (14.5%) a lobar bronchus was implanted into the main bronchus (Y-sleeve). Nine (10.8%) additional angioplastic procedures were performed.
Histology and staging
The histology was squamous cell carcinoma in 42 patients (50.6%), adenocarcinoma in 27 patients (32.5%), and neuroendocrine carcinoma grade II in 4 patients (4.8%). In 9 patients (10.8%) miscellaneous histologies were recorded (five large cell carcinomas, three undifferentiated carcinomas, one polymporphcellular carcinoma). In 1 patient after neoadjuvant chemotherapy no viable tumour tissue was present. Stage I was present in 29 patients (34.9%), stage II in 24 patients (28.9%), Stage IIIA in 21 patients (25.3.2%), IIIB in 4 patients (4.8%), and stage IV (after preoperative treatment of a solitary brain metastasis) in 3 patients (3.6%). In 1 patient no tumour stage was given. Mediastinal lymph nodes were positive in 21 patients (25.3%).
Causes of death
Forty patients (48.8%) died during the observation period 0 to 64 months after operation (mean 14.57, median 12.85, SD 13.36). Ninety-five percent of all deaths (n = 38) occurred within the first 3 postoperative years.
Twelve patients (14.4%) died tumor free. Six patients died due to postoperative complications (one pneumonia, one myocardial infarction, one mesenteric embolism, one gastrointestinal bleeding, one cerebral infarction, and one anastomotic dehiscence) representing an overall procedure related mortality of 5.5%. The details of these patients have been reported in a previous study [6]. Six patients (7.2%) died tumor free (one leucopenic sepsis during adjuvant chemotherapy, two pneumonias, one pulmonary embolism, and two accidents) 2 to 64 months postoperatively (mean 18.99, median 12.43, SD 21.62).
Five patients (6%) with tumor recurrence died of nontumor related causes (one myocardial infarction, one suicide, one stroke, two irradiation pneumonias). The follow-up period of surviving patients ranges from 5 to 100.7 months (mean 43.23, median 39.6, SD 27.96).
Tumor recurrences
Thirty-one patients (37.3%) developed tumor recurrence. Local and distant recurrence were observed in 20 patients (24.1%) each. In 11 patients (13.3%) only distant metastases were found, in 11 patients (13.3%) isolated local recurrence, and in 9 patients (10.8.8%) mixed recurrence.
The initial site of recurrence was local in 14 patients (16.8%) and distant in 16 patients (19.3%). Only in 1 patient (1.2%) local and distant recurrence were diagnosed simultaneously. Excluding nonradical resections disease-free interval was 1 to 42 months (mean 12, median 9, SD 11). In 13 patients (15.7%) a second manifestation of tumour recurrence on a different anatomic region was observed 1 to 48 months later (mean 11.08, median 6, SD 12.7).
Eighty-six percent of all recurrences were found within the first 2 postoperative years and 97% during the first 3 postoperative years, the latest recurrence being observed 42 months after primary resection. The details of TNM stage and recurrence are illustrated in Table 1. Recurrence localization is given in Tables 2 and 3.
|
|
|
Primary lymph node recurrence
Primary lymph node recurrence was seen in 10 patients (12%) 2 to 35 months (mean 16.9, median 18) postoperatively. Survival after diagnosis ranged from 1 to 19 months (mean 9.83, median 11). The bronchial tree showed signs of compression in 3 patients and invasion in another 5 patients (see below). Only 3 patients developed distant metastasis.
Secondary lymph node recurrence
Secondary lymph node recurrence was diagnosed 4, 4, 16, and 23 months postoperatively. The distance between primary distant recurrence and secondary local recurrence was 1, 2, 5, and 18 months. Patients survived 0, 1, 3, and 6 months after diagnosis of secondary lymph node recurrence. No progression to infiltration or compression of mediastinal structures was seen.
Intrabronchial recurrence
Intrabronchial recurrence was observed in 5 patients (6%). In 3 patients (including 1 patient with nonradical resection) this type of recurrence was preceded by mediastinal lymph node recurrence 1, 8, and 10 months earlier, in 2 patients both sites were involved simultaneously. No isolated intrabronchial recurrence was found. None of these patients developed distant recurrence.
Secondary oncologic procedures
Four patients (4.8%) underwent secondary oncological thoracic surgical procedures: one secondary sleeve pneumonectomy due to intrabronchial recurrence, one ipsilateral secondary lobectomy with thoracic wall resection due to intrapulmonary recurrence, two contralateral sleeve resections due to a metachronous tumour 53 months and a synchronous tumour 4 months postoperatively according to the criteria of Martini and Melamed [5]. The last 2 patients are alive and tumor free 17 and 74 months after the second intervention respectively.
Patterns of recurrence in detail
Distant recurrence
In 2 patients, who had been treated for singular brain metastasis preoperatively, it cannot be assessed whether their postoperative brain metastases represent local treatment failure or distant recurrence. Nevertheless they were included.
Distant recurrence was diagnosed in 11 patients (13.2%). The location is illustrated in Table 3. The first distant recurrence was found 1 to 42 months postoperatively (mean 10.64, median 9, SD 11.2), the second 2, 4, 12, and 48 months later.
Eight patients (9.6%) died 3 to 28 months postoperatively (mean 14.9, median 14, SD 8.6), 0 to 17 months (mean 7.55, median 7, SD 4.99) after the diagnosis of first recurrence and 1, 2, and 5 months after the diagnosis of second recurrence. All patients except one, who died of stroke, died of tumor-related causes.
At the end of the study 3 patients are alive 27, 49, and 70 months after operation. In 1 patient, who survives now 70 months, a singular brain metastasis had been resected 6 months after sleeve resection. Four years later a second sleeve resection was performed due to a contralateral metachronous tumor. He is alive 70 months after the primary intervention and 17 months after the second sleeve lobectomy.
Local recurrence
One patient developed colon carcinoma during the observation period and was operated. One month after abdominal surgery he developed thoracic wall recurrence. In another patient, who had been operated for colon carcinoma 2 years before lung resection, intraabdominal lymph node enlargement was observed simultaneously with multifocal local recurrence. In both patients the origin of intrathoracic tumour recurrence could not be surely assessed, however, we included them in the analysis. Nine local recurrences (10.8%) were observed 2 to 35 months (mean 17.3, median 12, SD 13.02) postoperatively, 8 patients (9.6%) died 4 to 34 months (mean 17.6, median 14.2, SD 11.7) postoperatively and 0 to 8 months (mean 2.13, median 1.5, SD 2.59) after diagnosis of recurrence.
Unifocal recurrence
Unifocal recurrence (n = 5) was found in the residual lung (n = 2), the mediastinal nodes (n = 1), in the mediastinum (n = 1), and in the thoracic wall (n = 1). Two patients died of nontumor-related causes: one of irradiation pneumonitis (in fact intrapulmonary local recurrence was only diagnosed at autopsy), the second patient had developed local recurrence in the ipsilateral lung and was reoperated (middle lobe lobectomy with thoracic wall resection). She also underwent adjuvant irradiation and died of irradiation pneumonitis 3 months after surgery.
In 2 patients local recurrence occurred 12 months postoperatively and was located in the mediastinum and the thoracic wall. They died within 1 month after diagnosis. One patient with isolated mediastinal lymph node recurrence was irradiated and is alive 41 months after primary operation and 6 months after diagnosis of recurrence.
Multifocal recurrence
In all 4 patients with multifocal recurrence mediastinal lymph nodes were involved (bifurcation, hilus, paraesophageal), and death occurred 1, 1, 2, and 8 months after recurrence. One patient with multifocal recurrence located intrabronchially and in the bifurcational nodes underwent secondary sleeve pneumonectomy. She developed mediastinal lymph node recurrence and liver and spleen metasases during adjuvant chemotherapy and died 8 months after resurgery.
The second recurrence involved hilus lymph nodes combined with stenosis of the main bronchus. The third patient presented with paraesophageal lymph node recurrence combined with lymphangiosis of the ipsilateral lung and pleural effusion. Tumor invasion of the main bronchus occurred 1 month after the diagnosis of recurrence. The fourth patient presented with enlarged upper and lower mediastinal nodes with bronchial invasion and died of myocardial infarction 1 month after recurrence diagnosis.
Mixed recurrence
Mixed recurrence was observed in 9 patients (10.8%) 2 to 21 months postoperatively (mean 8.33, median 5, SD 6.91). A second recurrence developed 1 to 18 months later (mean 8.67, median 6, SD 6.91). All patients died 4 to 41 months postoperatively (mean 17.83, median 16.27, SD 13.51), 1 to 21 months (mean 10.56, median 11, SD 7.38) after diagnosis of first recurrence and 0 to 6 months (mean 1.89, median 1, SD 1.89) after the diagnosis of second recurrence.
Synchronous mixed recurrence (n = 1)
This patient presented with lung metastases and enlarged lymph nodes in the upper mediastinum and thoracic wall recurrence 5 months after operation. Fourteen months later adrenal gland metastases occured and he died 19 months postoperatively.
Distant recurrence followed by local recurrence (n = 5)
Distant recurrence occured 2 to 11 months postoperatively (mean 5.6, median 5, SD, 3.58) and was located in the lung (n = 1), liver (n = 1), bone (n = 1), brain (n = 1), and adrenal gland (n = 1). Local recurrence occurred 1 to 18 months (mean 6.4, median 5, SD 1.8) after distant recurrence. In 4 patients mediastinal lymph nodes were involved. Additionally malignant pleural and pericardial effusion and thoracic wall recurrence were observed. All patients but one, who comitted suicide, died of tumor-related causes within 6 months after diagnosis of second recurrence.
Local recurrence followed by distant recurrence (n = 3)
Local recurrence occurred 3, 18, and 21 months (mean 24.67, median 30, SD 9.64) after primary resection and was located in the mediastinum in all 3 patients, resulting in stenosis of the main and intermediate bronchus in 2 patients.
Distant recurrence occurred 2, 12, and 18 months (mean 10.67, median 12, SD 8.08) after local recurrence in the lung (n = 1) and bones (n = 2). Survival after primary intervention was 5, 34, and 41 months (mean 26.44, median 33.93, SD 19.1), after primary recurrence 2, 15, and 19 months (mean 12, median 15, SD 8.89), and 0, 1, and 3 months (mean 1.33, median 1, SD 8.89) after secondary recurrence.
Nonradical resections
In 6 patients (7.2%) a nonradical resection was performed. One patient is lost to follow-up, 1 patient died without clinical evidence of disease 7 months postoperatively due to heart failure. One patient is alive without clinical evidence of disease 26 months after surgery. Two patients died of tumor recurrence involving mediastinal lymph nodes. In 1 patient bifurcational lymph node recurrence invading the bronchial tree with lethal hemoptysis occurred 12 months after surgery.
The second patient developed mediastinal lymph node recurrence 5 months after surgery followed by malignant pleural effusion 5 months later. In the 13th postoperative month invasion of the lower lobe bronchus occurred. He died 14 months after resection. The third patient showed local progression shortly after operation, developed pleural carcinosis and brain metastases after 12 months, and died 13 months after surgery.
Statistical evaluation
The following variables demonstrated no influence on recurrence rate: T stage, angioplasty, technique of bronchial reconstruction, localization of the tumor, neoadjuvant chemotherapy.
Increased recurrence rate was observed in patients with right-sided resections (chi-square, p = 0.003), in case of lymph node involvement (N0 vs. N1,2,3, p = 0.045) and in stage IIIA compard to stage I (p = 0.03). The N0 patients characterized a significantly lower recurrence rate than N2 patients (p = 0.041), squamous cell histology revealed significantly lower recurrence rate than nonsquamous cell histology (p = 0.029).
In a multiple logistic regression model only stage (p = 0.005, OR = 1.982, Cl = 1.233–3.186) and right side (p = 0.014, OR = 0.177, CI = 0.045–0.704) remained significant predictors. The quality of the model (R2 = 32) was unsatisfactory to predict recurrence.
|
Comment |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentReferences |
|---|
With only one patient missing we were able to obtain nearly a complete follow-up. Our recurrence rate lies within the range of the literature [10–14] (Table 4). Tumor recurrence is reported usually within the first 3 postoperative years with the majority of patients dying during the first 2 postoperative years [15], which is comparable to our data. Nevertheless, because tumor recurrence and second malignancy may also occur at a later time, routine follow-up should be continued at least 5 years.
|
|
If our patterns of recurrence are compared with those after conventional resections there is a striking difference in the rate of mixed recurrence, which is twice as high in our series. Probably the other authors only classify synchronous recurrence as mixed recurrence while metachronous recurrences are classified according to the first site of recurrence. This is a correct interpretation but simply ignores the natural course of recurrent disease.
Our data are derived from radiologic and endoscopic investigations. Clinical tumor staging revealed no other recurrence sites than those reported. It must be taken into account that our patients had not undergone autopsy and, therefore, the exact pattern of recurrence remains unknown.
Nevertheless, we identified sites of recurrence leading to severe clinical symptoms and death while others remained asymptomatic in different subsets of patients. This observation suggests different biological behavior depending on the site and pattern of recurrence.
Nearly one third of all patients with tumor recurrence exclusively developed distant metastases without detectable local recurrence at all. Either local microscopic disease remained stable, progressed extremely slowly or local radicality had been achieved by operation and distant metastases had developed before surgery. In this group a high potential of distant spread and growth in combination with low local aggressiveness is imaginable. Another third presented with very aggressive local recurrence without clinical development of distant metastases. In all but two of these patients histologic local radicality had been achieved during primary resection. Invasion and/or compression of the tracheobronchial tree due to enlarged mediastinal lymph nodes occurred in 7 of 11 patients suggesting the biological capability of destruction of local anatomic structures and quick lymphatic tumor spread. Interestingly only 3 of 10 patients with primary mediastinal lymph node recurrence developed distant metastasis and no patient with mediastinal and intrabronchial recurrence. Either the potential of hematogenous spread was missing or tumor growth in extrathoracic sites was extremely slow.
In 4 of 5 patients with mixed recurrence beginning with distant metastasis mediastinal lymph node involvement indicated no aggressive behavior. One reason might be that these patients had received chemotherapy. More intensive clinical observation could also have led to earlier diagnosis of lymph node enlargement. Another explanation might be that these patients simply did not live long enough to develop invasion or compression of the bronchial tree. As well the biological potential of local destruction could have been low in contrary to the capability of hematogenous spread and extrathoracic tumor growth. Only 4 patients revealed an equivalent local and distant aggressiveness.
Conclusion
We conclude that intrabronchial recurrence after radical bronchial sleeve resection is a result of regional lymph node recurrence. If such a patient undergoes secondary resection adjuvant mediastinal irradiation should be considered. The existence of an isolated intrabronchial recurrence after radical resection is questionable. The pattern and natural history of recurrence is a result of different genetic characteristics of the primary tumor. The concept of surgical radicality and histopathologic staging fails to predict or explain recurrence. However, our number of patients is low and more descriptive studies are needed to collect data on the natural history of postoperative tumor recurrence. The investigation of the relationship between genetic changes in tumor cells and the clinical recurrence patterns should allow us to define risk groups that might profit from adjuvant chemotherapy and/or irradiation in the future.
|
References |
|---|
|
TopAbstractIntroductionPatients and methodsResultsCommentReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
