| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Pneumologie der 1. Medizinischen Klinik und Poliklinik (Drs. Riedel and Lembeck) and Chirurgische Klinik und Poliklinik (Drs. Stein, Siewert, and Mounyam), Klinikum rechts der Isar, Technische Universität, München, Germany.
Correspondence to: Martin Riedel, MD, 1. Medizinischen Klinik und Poliklinik, Klinikum rechts der Isar, Ismaninger Strasse 22, D-81675 München, Germany; e-mail: m.riedel{at}dhm.mhn.de
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Study objective: To evaluate the utility of bronchoscopy in the preoperative assessment of airway invasion by supracarinal esophageal cancer.
Design: We prospectively analyzed 220 bronchoscopies in 166 consecutive patients with supracarinal esophageal cancer and correlated the findings with operative results and survival.
Results: In 126 bronchoscopies (57.3%), no abnormal findings could be seen in the airways. Compared with histologic and cytologic results, the normal macroscopic appearance of the airways had a negative predictive value of 94.4%, but the positive predictive value of all macroscopic abnormalities for the diagnosis of airway invasion was low, particularly after radiation therapy. Endoluminal tumor mass, protrusion of the posterior tracheal wall, and signs of mucosal invasion were visible in 5.9%, 28.6%, and 4.1% of the bronchoscopies, respectively. However, in only 8.6% of the 220 bronchoscopies, cancer invasion was proved by biopsy or cytology. Bronchoscopy with biopsies and brush and washing cytology examinations was the sole decisive staging procedure, enabling the exclusion from surgery because of airway invasion in 18.1% of otherwise potentially operable patients, with an overall accuracy of 93.3% (95% confidence interval, 86.7 to 97.3%). The results of bronchoscopy were falsely negative in six patients, who all underwent surgery after neoadjuvant therapy.
Conclusions: Fiberoptic bronchoscopy with systematic multiple biopsies and brush and washing cytology examinations is an accurate procedure in evaluating the possible invasion of supracarinal esophageal cancer into the airways. Macroscopic findings alone are not reliable; errors in sole bronchoscopic inspection would have resulted in operations that would be unlikely to help the patients or would have inappropriately excluded patients from surgery.
Key Words: bronchoscopy • esophageal neoplasms • esophagectomy • staging • trachea
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our previous study2 in 116 patients (with both supracarinal and infracarinal cancer) showed a high accuracy (95.8%) of bronchoscopy when used with biopsies and brush cytologic examinations in proving or excluding airway invasion of the cancer in potentially operable patients. Macroscopic findings alone were not reliable. Bronchoscopy used with biopsies and brush cytologic examinations enabled the decision to exclude the patient from surgery because of airway invasion in 9.7% of patients with otherwise potentially operable conditions.2 The aim of the present study was to confirm these results prospectively in a new patient population with supracarinal esophageal cancer only and, in addition, to evaluate the usefulness of routine tracheal brushings and washings from both lungs for the detection of airway invasion of the esophageal cancer.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
After obtaining informed consent, fiberoptic bronchoscopy was performed by the pulmonologists, as previously reported.2 11 Washings for cytologic examination were obtained by suction from both lungs. The complete tracheobronchial tree was examined; special attention was paid to the trachea and main bronchi. All direct tumor signs (ie, exophytic intraluminal growth, infiltration of the tracheobronchial wall, or esophagotracheal fistula) and indirect tumor signs (ie, distortion or compression of normal structures, disappearance of mucosal corrugations, altered structure of the mucosa, erythema with teleangiectatic blood vessels, protrusion at the posterior wall of the trachea or a major bronchus, widened bifurcation, and rigid and fixed pars membranacea during breathing or coughing maneuvers) were recorded, and three to five forceps biopsy specimens as well as brush cytology samples were taken from these areas. If no abnormalities could be seen at bronchoscopy, brush cytology specimens and three to four biopsy specimens were routinely taken from the pars membranacea of the right main bronchus, the left main bronchus, and the distal, medial, and proximal part of the trachea, respectively. The quality of the biopsy specimens, as assessed by the pathologist, was considered optimal when at least two pieces > 2 mm were obtained from the region of interest and was considered adequate when at least two specimens > 1 mm were obtained. The brush and washing samples for cytology were considered adequate when at least four cell-rich slides, consisting predominantly of columnar epithelial cells, could be analyzed microscopically.
After the initial staging, patients with potentially operable conditions proceeded directly to esophageal resection and those with locally advanced disease (defined as stages T3 or T4)12 received neoadjuvant therapy with 30-Gy radiation and 5-fluorouracil. This rather mild regimen was chosen in order to minimize the radiochemotherapy-associated postoperative morbidity and to allow outpatient treatment. Patients showing an incomplete response to radiochemotherapy underwent a second restaging bronchoscopy. In the study period, no patient was operated on without undergoing at least one preoperative bronchoscopy. The resectability criteria were established prior to the commencement of the study and included the exclusion of distant metastases and the exclusion of airway invasion, determined by bronchoscopy with cancer-positive microscopic examination. Subtotal transthoracic en bloc esophagectomy with two-field lymphadenectomy was the procedure of choice. Reconstruction was performed with a gastric tube anastomosed to the cervical esophagus. Tumor extent and airway invasion were assessed intraoperatively and by histopathologic examination of the resection specimens. Pathologic stage was determined on the basis of the Union Internationale Contre le Cancer TNM classification guidelines.12
Patients who were operated on were followed-up every 3 months for the first year and at 6-month intervals thereafter. Episodes suggestive of cancer invasion into the airways were specifically looked for. Patient death records were obtained from the hospital or through our tumor registry. The fate of all patients was ascertained by June 30, 1999, through telephone calls with the patients, their relatives, or their primary physicians.
For the therapeutic decision, the results of bronchoscopy were reported and the patients classified into the following three groups: (1) no invasion (ie, no macroscopic abnormality, and no malignancy on routine biopsy specimens or washing and brush cytology); (2) possible invasion (ie, macroscopic abnormality without microscopic proof of malignancy from samples taken from the macroscopically suspect areas as well as from routine biopsy specimens or washings and brush cytology); (3) proven invasion (ie, histologic or cytologic proof of malignancy, with tumor type identical to the type of esophageal cancer). For the assessment of the validity of bronchoscopy in the diagnosis of tracheobronchial invasion, an R0 resection (ie, complete tumor resection12 ) with regard to the airways and with survival of > 6 months without any pulmonary problems was considered as the "gold standard" for the absence of airway invasion.
The variables are expressed as the mean ± SD or as a percentage of
the group they were derived from. Statistical differences between
groups were determined by Student’s t test or the
2 test, where appropriate. All statistical
tests were performed using a 5% level of significance. Standard
methods for proportions were used. The binomial distribution was used
to compute the 95% confidence intervals (CIs) for the frequency of
bronchoscopic abnormalities and for their validity indexes. The
sensitivity, specificity, and accuracy of the assessment were defined
as the proportion of correctly assessed patients in those with proven
involvement, in those without proven involvement, and in all patients,
respectively.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Initial Bronchoscopy
Table 2
lists the macroscopic findings in the trachea and main bronchi at the
first bronchoscopy and their correlations with histology and cytology.
Only 14 patients showed a direct tumor sign. Of the indirect tumor
signs, localized mobile protrusion of the pars membranacea
was the most frequent, occurring in one fourth of the patients. In
54.2% of patients, the airways were macroscopically normal.
Macroscopic findings of an endoluminal mass, signs of mucosal
infiltration, and rigid protrusion of the posterior tracheal wall best
predicted microscopic involvement.
|
|
Clinical Course After the Initial Bronchoscopy
Figure 1
shows the outcome of the patients after the first bronchoscopy.
Thirty-four patients proceeded directly to esophageal resection, on
average 9.9 ± 7.2 days after undergoing the bronchoscopy. Nineteen
of the 79 patients who ultimately were treated palliatively were
excluded from surgical resection because of airway invasion diagnosed
by bronchoscopy (biopsy specimen confirmation, 12 patients; brush
cytology confirmation, 7 patients; and washing confirmation, 7
patients). In these 19 patients, bronchoscopy was decisive for the
therapeutic decision. Of the other 60 palliatively treated patients, 19
were rejected from surgery because of an extensive local tumor mass, 35
were rejected due to general inoperability, 11 were rejected due to
distant metastases, and 5 were rejected for other reasons (more than
one reason per patient was possible). Of these 60 patients, 34 had
normal bronchoscopic findings and 26 had some macroscopic abnormalities
with normal microscopy in samples taken from these areas.
|
Restaging Bronchoscopy After Neoadjuvant Therapy
Table 4
lists the macroscopic findings in the trachea and main bronchi at the
restaging bronchoscopy and their correlations with histology and
cytology. One patient with normal results of a macroscopic examination
had results of a routine brush cytology that were strongly suspicious
of cancer at the restaging. This patient underwent an R0
resection and is well 18 months after surgery. The cytologic diagnosis
of airway invasion at bronchoscopy was, thus, falsely positive in this
patient.
|
In 4 of the 35 patients with both normal macroscopic and microscopic examination at the restaging bronchoscopy, airway cancer invasion was diagnosed at surgery and was confirmed by the histologic examination of the resectate. Thus, the second bronchoscopy was falsely negative in these four patients.
Taking bronchoscopic biopsy as the "gold standard" (Table 3) , the positive predictive value of all macroscopic abnormalities at the second bronchoscopy was even lower than at the initial bronchoscopy; only 1 of 18 histologic samples taken from endoluminal masses and from areas suggestive of tumor invasion showed cancer cells. By combining all histologic and cytologic examinations and taking them as standard, the positive predictive value of all macroscopic abnormalities at the restaging was only 5.6%, but the negative predictive value was 97.2%.
Esophageal Resection
Eighty-five of the 86 patients who underwent esophagectomy
underwent surgery with curative intent. The patient with normal
results of macroscopic examination but results of routine brush
cytology that were strongly suspicious of cancer at the second
bronchoscopy was operated on with palliative intent but underwent an R0
resection. The hospital mortality rate was 5.8% (5 of 86 patients). An
R0 resection was achieved in 63 of 86 patients (73.3%) (Fig 2
). In 23 patients, complete tumor resection could not be achieved due to
airway invasion in six patients and invasion into other structures in
the remainder of the patients. All of the six patients with airway
invasion received neoadjuvant therapy before surgery, and five of them
underwent a restaging bronchoscopy. Five of the six patients with
airway invasion at surgery had no macroscopic abnormalities at the last
bronchoscopy (performed 21.6 ± 25.6 days before the operation
[range, 2 to 69 days]); one patient showed mobile protrusion 2 days
before surgery. Routine histology and cytology samples had been
negative in all six patients.
|
|
|
Of the 80 patients without tracheal invasion at surgery, 44 were alive 17.7 ± 9.1 months after resection (range, 6 to 33 months) without pulmonary problems. Thirty-six patients (including the 5 not discharged from the hospital) died an average of 11.7 ± 8.7 months after the resection (median, 10.5 months). Of these 36 patients, 12 died of causes unrelated to the airways, 6 patients had airway invasion diagnosed shortly before death (11.5, 12, 12, 15, 30, and 32 months after surgery), and in the other 18 patients airway invasion was unlikely but could not be definitively excluded.
Fourteen of the 19 patients who did not undergo surgery but who did have airway invasion diagnosed at bronchoscopy died 4.8 ± 4.4 months (median, 3 months) after the bronchoscopy and 5 patients survived an average of 14 ± 9.5 months (median, 19 months). Of the 61 palliatively treated patients who had not experienced airway invasion at bronchoscopy, 38 died 7 ± 5.7 months (median, 6.3 months) after the bronchoscopy. Of these 38 patients, 9 died of causes unrelated to the airways, 2 had airway invasion diagnosed shortly before death (2.5 and 18 months after the bronchoscopy), and in the other 27 airway invasion could be neither proved nor excluded. Twenty-three patients survived 17.8 ± 4.2 months (median, 15 months) after the bronchoscopy. Two of these 23 patients developed an airway invasion 12 months after the bronchoscopy, 13 survived without pulmonary problems, and airway invasion could be neither proved nor excluded in the other 8 patients.
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our study showed a high overall accuracy (93.3%) of bronchoscopy with biopsy and brush and washings cytology in proving or excluding airway invasion of supracarinal esophageal cancer in patients with potentially operable conditions. Bronchoscopy with biopsy and cytologic examination identified airway invasion that was not demonstrated by other staging procedures. Bronchoscopy was the sole decisive staging procedure, enabling exclusion from surgery because of airway invasion in 11.4% of the overall study population and in 18.1% of patients with otherwise potentially operable conditions. For these patients, bronchoscopy clearly altered the course of therapy, as they otherwise would have undergone esophagectomy in the presence of tracheobronchial tumor invasion.
The total incidence of airway invasion of esophageal cancer in this study (25 patients [ie, 15.1% of the total 166 patients]) was comparable to that previously reported.2 3 4 5 Bronchoscopy identified 19 of these 25 patients; in 6 cases the bronchoscopy was falsely negative.
Suspect macroscopic abnormalities in the central airways were found in 45.8% of our patients at the initial bronchoscopy. However, microscopic proof of malignancy through histology could be obtained in 14.1% of these abnormalities only. Brush cytology was positive for cancer in only 9% of abnormalities, and washing cytology in only 4.9% of the macroscopic abnormalities. Eighteen patients with macroscopic abnormalities but without microscopic proof of malignancy eventually underwent an R0 resection; they would have been rejected for curative surgery if the diagnosis of airway invasion had been based on macroscopic findings only. The assessment of indirect tumor signs is subjective; therefore, the diagnosis of airway invasion must be based on microscopic proof of cancer from the suspected areas. All three kinds of specimens (biopsies, brushings, and washings) should be taken, because none of these examinations alone was positive in all cases.
Protrusion of the posterior tracheal wall with normal motility during breathing and coughing may only be due to the presence of a bulky tumor in the vicinity and does not preclude radical resection.2 3 4 8 9 None of the biopsy specimens or brushings from these areas both at the initial and the restaging bronchoscopies (Tables 2 and 4) was positive for cancer. Further studies should address the question of whether patients with this isolated finding should proceed directly to surgery or undergo neoadjuvant therapy with the aim of reducing the tumor mass preoperatively.
It is difficult to assess from our results the value of routine biopsies from macroscopically normal areas. Only one of the biopsy specimens taken in patients with no macroscopic abnormality was positive for cancer. The taking of biopsy specimens from several regions in the absence of visible abnormalities represents more sampling than in the usual setting, and even the few biopsy specimens that would be positive might be missed under less rigorous testing conditions. Thus, the applicability of these data to the usual clinical setting may be limited. The benefit of routine biopsies might be higher if specimens were taken only from areas nearest to the esophageal tumor, as located by CT scanning. However, taking samples for cytology by washings and tracheal brushings is useful even if there is no visual evidence of mucosal changes, as documented in five of our patients at the initial bronchoscopy.
At the restaging bronchoscopy after neoadjuvant radiochemotherapy, we found direct tumor signs in eight patients (14.8%), with microscopic proof of cancer in only one patient. Macroscopic abnormalities after radiation therefore must be interpreted even more carefully than at the initial staging, because a white soft intraluminal mass and altered mucosal structure could result from radiation-induced peritumorous inflammatory or fibrous reactions alone and not from tumor invasion. Interestingly, all six patients with falsely negative bronchoscopies underwent neoadjuvant therapy before the operation. In none of the 34 patients who proceeded directly to surgery was the initial bronchoscopy falsely negative, probably because in our protocol only patients with early tumors are considered to be suitable for primary resection. Therefore, both overstaging in the evaluation of direct tumor signs without histologic proof and understaging are more common after neoadjuvant therapy than at the initial staging.
The assessment of long-term survival was not the aim of this study. However, a postoperative survival of > 6 months without any pulmonary problems is a further proof of an R0 resection of the airways, strengthening the "gold standard" with which our bronchoscopic findings were correlated. It is impossible to ascertain whether the exclusion of airway invasion by bronchoscopy was correct in the patients treated palliatively for other reasons. The natural tumor growth as well as the palliative therapy in these patients could have influenced tumor invasion in the airways, making any later evaluation of the accuracy of the bronchoscopy useless. Nevertheless, the better survival rate of the patients who did not undergo surgery and did not have airway invasion at the initial staging, compared to those who did have airway invasion, suggests that the exclusion of airway invasion at that time was probably correct in the majority of the patients.
In this study, 36.1% of patients (60 of 166 patients) were excluded from resection because of systemic metastases, extensive local tumor mass, or poor physiologic status documented only after bronchoscopy was performed. If this information had been available prior to bronchoscopy, the procedure could have been avoided. We believe that bronchoscopy should be performed as the last investigation in the staging workup, after inoperability has been excluded by other obligatory staging examinations. The results of bronchoscopy in patients with inoperable conditions without pulmonary symptoms will not substantially modify their palliative treatment.
Interestingly, most of our patients did not have distant metastases despite having stage T4 disease status and the significant size of the primary tumor (Table 1) . This is in concordance with reports8 13 on patients with malignant esophagotracheal fistulas, in which only half of the patients had metastatic disease. Possibly, tumors presenting with airway invasion may have a more marked propensity toward local spread rather than early distant metastases, which occur only at a later stage in the life cycle of such tumors. Untreated, most patients will die of the sequelae of local tumor invasion rather than the ravages of disseminated disease.
The concept of this study was intentionally similar to our previous investigation,2 because we considered it appropriate and justified to confirm the former results in a new population. This study confirmed and expanded the results of the previous study. The addition of routine tracheal brushings and washings for cytology in all patients, and perhaps more experience and rigorous sampling from all macroscopic abnormalities, led to more exclusions of patients from surgery because of airway invasion in this study (11.4% vs 6% of all patients; and 18.1% vs 10% of the potentially operable patients in this study vs the previous study2 ). The overall accuracy of bronchoscopy in proving or excluding tracheobronchial invasion in patients with otherwise potentially operable conditions was comparable in both studies.
Because the tumor is situated in a place that is distant from the central airways in patients with infracarinal esophageal cancer, direct airway invasion by the tumor is extremely rare in them. Such patients present other problems regarding the preoperative bronchoscopic evaluation (such as the infiltration of the lower lung lobes from the outside), and they were, therefore, excluded from this analysis. They are described in a separate report.11
In conclusion, this study showed a high accuracy (93.3%) of bronchoscopy with biopsy and brush and washings cytology in proving or excluding airway invasion of supracarinal esophageal cancer in patients with potentially operable conditions. Macroscopic findings alone were not reliable. Bronchoscopy with biopsy and brush and washing cytology was the sole decisive staging procedure, enabling exclusion from surgery because of airway invasion in 18.1% of patients with otherwise potentially operable conditions. The examination should be performed as the last investigation in the preoperative workup.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Received for publication July 10, 2000. Accepted for publication January 9, 2001.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Baisi, L. Bonavina, and M. Riedel Preoperative Bronchoscopic Biopsies and Staging Chest, June 1, 2002; 121(6): 2081 - 2082. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
