| 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 Departments of Pulmonology (Drs. Pasic, Postmus, and Sutedja), Pathology (Drs. Grünberg and Mooi), and Surgery (Dr. Paul), Vrije Universiteit Medical Center, Amsterdam, the Netherlands.
Correspondence to: Thomas G. Sutedja, MD, PhD, Department of Pulmonology, Vrije Universiteit Medical Center, PO Box 7057, 1007 MB Amsterdam, the Netherlands; e-mail: tg.sutedja{at}vumc.nl
Abstract
Study objectives: Microscopic residual disease in the bronchial resection margins after surgical resection of lung cancer is rare, and its clinical significance remains unsettled. We studied the natural history of patients with carcinoma in situ (CIS) at their bronchial resection margins to focus on the issue of stump recurrence.
Methods: Eleven individuals who had undergone radical surgery for N0M0 lung tumors were found to have CIS at the bronchial resection margins. All of the resection specimens were reviewed with respect to the pattern of CIS extension and reclassified as follows: superficial CIS, involving surface epithelium only (CIS-S), CIS extending into the submucosal gland ducts but not deeper (CIS-D), and CIS extending into submucosal gland acini (CIS-A). Patients were followed using autofluorescence bronchoscopy and high-resolution computer tomography. Clinical parameters and the local extent of CIS at histology review were correlated with outcome.
Results: Median follow-up was 35 months (range, 15 to 89). Histology review showed two CIS-S cases, six CIS-D cases, and three CIS-A cases. All of the patients with CIS-A developed stump recurrences in contrast with those with only CIS-S. Three patients with CIS-D have developed metachronous primaries in the contralateral lung, whereas the stump region remained free of tumor.
Conclusions: The presence of CIS in the bronchial resection margin after resection of lung cancers is associated with stump recurrences. Although absolute numbers are too small for firm conclusions, our data suggest that those with deep glandular extension of CIS bear the highest risk of early recurrence. However, the development of new primaries away from the stump region and the possible development of distant disease are equally relevant considerations with respect to the choice of additional therapy.
Key Words: carcinoma in situ • lung cancer • residual disease • surgery
When mediastinal lymph nodes are free of metastasis, radical resection offers the best chance of a cure for lung cancer. Even when the bronchial resection margin appears free of disease at a frozen section, the definitive histologic investigation reveals microscopic residual disease (MRD) at the resection margin in a small number of cases. The clinical significance of this finding with regard to stump recurrence has remained unsettled, and there are only a few reports on this issue.12345678910111213 Superficial mucosal or intramucosal tumor spread, in contrast with more extensive infiltration, that is, infiltration involving extramucosal, peribronchial tissue and the lymphatics,41012 seems to have a favorable outcome. The majority of studies published so far have included patients with various TNM stages, and, in many cases, adjuvant therapy was instituted after the MRD had been diagnosed. The inclusion of patients with deeper extension beneath the mucosa together with patients with nodal diseases hampers the judgment of the prognostic impact of MRD. Also, differences in the methodology have precluded comparisons of the data, thereby precluding unequivocal conclusions as to the possible benefit of adjuvant treatment.
We, therefore, limited our study group to individuals with N0M0 resected non-small cell lung cancers, who were found to have carcinoma in situ (CIS) involving the bronchial resection margin at the definite histologic investigation. Additional treatment was not given unless there was histologic proof of the tumor recurrence. Patients were closely followed with autofluorescence bronchoscopy (AFB) and high-resolution computer tomography in order to study the natural history and outcome of patients with CIS. Data from the literature have also been reviewed.
Materials and Methods
All of the individuals who were referred to our department with CIS involving the bronchial resection margin, as evident from definitive histologic investigation, after their frozen section has initially been considered negative, were studied. During the surgery, formal lymph node dissection was performed in all of the patients, and only those with psN0M0 (Pathological Stage NOMO) lung cancer have been included in this final analysis. The patients with invasive carcinoma at the resection margin and follow-up of < 1 year postsurgery were excluded.
The main reason for referral to our center was the availability of, and experience with, AFB using an endoscope system (Laser Induced Fluorescence Endoscope; XillixR; Richmond, BC, Canada). This technique has been shown to be more sensitive for the detection of early changes in the bronchial mucosa.1415 Also, patients were referred because of our expertise in interventional pulmonology.1617
The histology of all of the resection specimens was reviewed by two pathologists (K.G. and W.M.) using World Health Organization criteria for classification.1819 In addition, the diagnosis of CIS was refined into the following three categories (Fig 1 ): (1) CIS limited only to the surface epithelium; (2) CIS involving surface epithelium and extending into the bronchial gland ducts but not deeper (CIS-D); and (3) CIS involving surface epithelium and extending into the mucosal gland ducts, as well as into the glandular acini.
|
The patients were regularly followed with AFB, as well as white light bronchoscopy to inspect the bronchial mucosa, especially the stump. Suspicious lesions were visually scored and biopsied. High-resolution computer tomography was performed at baseline and repeated at least once a year in order to evaluate extraluminal growth in the stump region and for the early detection of possible new lung primaries.
The detection of disease in the stump region was considered a local recurrence. The development of malignancy in the ipsilateral lung but in a different lobe or in the contralateral lung was considered distant disease.
Results
From June 1997 to November 2003, 13 individuals with CIS diagnosed in their bronchial resection margin were referred to our hospital for additional evaluation and follow-up. Patient characteristics and various clinicopathologic parameters are shown in Table 1 . The stained sections were reviewed and classified as described. At review, two of the cases were found to harbor invasive carcinoma at the bronchial resection margins and were excluded from additional analysis. The patterns of CIS extension are presented in Table 1, and examples are given in Figure 1.
|
Three of the six patients with CIS involving the surface epithelium and extending into the bronchial gland ducts but not deeper (CIS-D) have developed new primary carcinomas in the contralateral lung at 11, 19, and 28 months after surgery, without evidence of a stump recurrence. One of them subsequently received intraluminal treatment and external radiotherapy but died 4 years postsurgery, and another received palliative irradiation and died 30 months after surgery. Both deaths were attributable to the new primary carcinomas. The remaining patient planned to receive palliative treatment but died of cerebral vascular attack 21 months postsurgery. Two patients with CIS limited only to the surface epithelium are free of disease 15 and 25 months after surgery.
Discussion
Our data show that the risk of local recurrence after resection for long carcinoma depends not only on the presence of CIS in the resection margin but also on the presence of gland involvement. The risk for early recurrence appears to be the highest in the patients who have deep glandular involvement of CIS. The data also show that there is a concurrent high risk of developing new primary carcinomas. Both risks should be weighed in therapeutic decision-making.
Several groups have studied the significance of MRD at the bronchial resection margins, and various classification systems of MRD have been proposed.12345678910111213 Generally, the following two main patterns could be recognized: mucosal/intramucosal MRD and extramucosal MRD as firstly described by Cotton.1 Later studies on MRD have used a more refined classification, because there might be a difference between mucosal spread of CIS and the usually more dominant feature of deeper mucosal/extramucosal spread, which may extend to peribronchial tissue and involve submucosal structures.451012 The 5-year survival rates seem quite favorable only in the group with mucosal CIS, even if compared with patients after radical resection.10 However, the various criteria of MRD with respect to glandular involvement, together with the inclusion of patients with nodal diseases, and the application of additional treatments for the stump region make a proper judgment about the impact of MRD on outcome difficult.
So far, because of the retrospective nature of many studies and the lack of regular prospective examination with regard to the stump (without local treatment, such as radiotherapy) and distant region, no clear picture of disease recurrence and development could be given, especially regarding the stump itself. Therefore, it remains unclear whether adjuvant therapy may improve the outcome, especially in the case of CIS. Our study is unique in that we included only N0M0 patients with CIS-positive margins and followed them without additional interventions until a local and/or nonlocal recurrence developed.
The number of individuals in our series precludes firm conclusions. Reported percentages of residual disease at the resection margins have ranged from 1.2 to 14.7%.413 For the subgroup with CIS only, these numbers are even smaller, with largest study12 reporting 20 patients, only 13 of whom had N0 disease (Table 2 ). Lacasse et al20 have recently addressed the issue of incomplete resection in non-small cell lung cancer. Their report contains only 3 cases with CIS at the resection margin after radical resection in 399 patients. In view of this very low incidence, we believe that our series of well-characterized patients (CIS, N0, no local additional treatment, and close follow-up) do give an indication of the natural course of CIS, which may contribute to resolving the issue of the potential benefit of local treatment.
|
In the group with CIS-D, three cases have developed new malignancies but not stump recurrence. This fits with the hypothesis of field carcinogenesis and the presence of multiple premalignant/malignant cell clones, each progressing according to their own pace toward metachronous cancers.1221 It is clear that the stump will not always be the first to cause carcinoma recurrence in the presence of competing cell clones that are present in the entire mucosa.
In five of the six cases of recurrent disease, the tumor was detected early by autofluorescence bronchoscopy, indicating the superior sensitivity of this new imaging technique in detecting recurrence and new primary carcinomas,141516 which allows for earlier intervention and for the potential improvement in life expectancy.1617
The data presented in this study do not allow conclusions regarding the best course of action if CIS is detected preoperatively at frozen-section investigation. All of the cases entered in this study concerned bronchial margins that had been underdiagnosed (ie, inappropriately considered to be free of tumor) at frozen-section evaluation. It is generally known that preoperative frozen-section diagnosis has limited sensitivity. Handling of the unfixed specimen is less optimal, so that the frozen section may consist of pieces instead of a continuous and complete circumference of the bronchus, and the surface epithelial lining may be incomplete and damaged. In addition, there is considerable time pressure. Under these circumstances, small or even minute abnormalities, such as focal CIS, may escape detection at the frozen section, to become evident only at the final paraffin-section-based investigation. It seems reasonable to assume that this happens more often if the lesion is very small.
So far, our data show that the mucosal and superficial spread of CIS has not led to stump recurrence (zero of eight cases). Three patients have developed new primary carcinomas in the field, that is, metachronous cancer, a pattern that fits in with the data reported by Massard et al.12 Apparently, the risk of developing new primary carcinomas outweighs the risk of local recurrence in these patients. This may be explained by the long lead time for squamous cell carcinomas to develop. This finding puts in perspective the value of local radiotherapy in these patients. Also, given the current abilities for early detection in combination with the potentials of minimally invasive treatment modalities other than reamputation surgery, per se, a vigilant approach during follow-up is warranted. Only in the case when the review of histology shows deep tumor extension and/or invasive cancer and, of course, if frozen section analysis of the bronchial stump before leaving the operating room identifies CIS in the margins, should immediate adjuvant treatment be indicated.
In conclusion, a detailed histologic description of the pattern of spread of CIS in cases with initially radical resection provides prognostic clues as to the risk of local recurrence vs the risk of distant disease. Despite the small number of patients and the relatively short follow-up considering the lead time of carcinogenesis, our data do suggest that the risk for early stump recurrences is highest in the subgroup of patients with deep glandular CIS involvement. This, together with the field effect for developing new subsequent primaries and the threat of distant disease in the various TNM patient categories, puts into perspective the impact of MRD, which is not only the risk of a stump recurrence. Our data suggest that CIS in the stump is the expression of diffuse mucosal abnormalities, and, as such, is also a marker for the increased risk of metachronous cancers in the lung field, distant from the bronchial stump. In addition, a vigilant approach toward field carcinogenesis and the current ability to detect early new primary carcinomas is warranted.
Footnotes
Abbreviations: AFB = autofluorescence bronchoscopy; CIS = carcinoma in situ; CIS-D = carcinoma in situ involving surface epithelium and extending into the bronchial gland ducts but not deeper; MRD = microscopic residual disease
Received for publication December 3, 2004. Accepted for publication February 3, 2005.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
