| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
Women’s Imaging Center, Tristán Associates, 4518 Union Deposit Road, Harrisburg, PA 17111, e-mail: mguenin@tristans.com
Editor:
In a recent article (1) in the September 2000 issue of Radiology in which high-risk lesions diagnosed at stereotactic breast biopsy were assessed, Dr Philpotts and colleagues advocate wide local excision in all cases where lobular carcinoma in situ (LCIS) is found histologically, stating that "lesions diagnosed as LCIS, atypical, or suspicious for malignancy and those with unusual associated histologic findings, pathologist’s uncertainty, or mammographic-histologic discordance should be excised." Their recommendation is based on their experience with one patient. I do not believe such a conclusion is warranted.
Perhaps the most relevant work addressing this topic was performed by Liberman et al (2). On the basis of their experience in 14 patients with LCIS evaluated at stereotactic biopsy who subsequently underwent excision, they recommended excision of LCIS under more limited circumstances, namely, "when the percutaneous biopsy histologic features overlapped with those of DCIS [ductal carcinoma in situ], when a high-risk lesion was present, or when there was imaging-histologic discordance." They went on to state explicitly, "LCIS without these factors was not shown to require surgical excision. ..."
Dr Philpotts and colleagues seem to have taken the criteria of Liberman et al for the excision of LCIS under limited circumstances and broadened them, without support from their own data, to include all cases of LCIS. Their single case where LCIS was found at stereotactic biopsy yielded DCIS at surgery would have been triaged appropriately with the imaging-histologic discordance criterion of Liberman et al, since that single patient had an associated mass unexplained by a diagnosis of LCIS alone.
Perhaps Dr Philpotts and colleagues shared the reservations expressed by Gabriel (3) about basing management decisions concerning LCIS on data from such a small group of patients. If so, they should have stated explicitly in their article that they disagreed with the criteria by Liberman et al on the basis of personal preference. Instead, the reader is left to infer that Dr Philpotts and colleagues have proved the merits of universal excision of LCIS, which they have not.
The approach of Liberman et al makes inherent sense to me. Since LCIS has no palpable or imaging correlate, it has long been recognized as an incidental finding at histologic examination (4,5). As an example, if biopsy is performed to evaluate microcalcifications, if those microcalcifications are completely removed in the course of vacuum-assisted directional biopsy, and if histologic analysis reveals calcification in a small degenerated fibroadenoma with LCIS in the adjacent breast tissue, then the work-up is complete. LCIS is a serendipitous finding. We are no more obligated to recommend wide local excision in this case than a surgeon would be obligated to reexcise for clean margins if LCIS were discovered incidentally during a similar diagnostic surgical biopsy (5).
In summary, Liberman et al have developed a reasonable set of criteria guiding the recommendation for excision (or against excision) when LCIS is diagnosed at stereotactic biopsy on the basis of a thorough analysis of the known natural history of LCIS. While Liberman et al cautioned that their series was small and that a larger study was needed, Dr Philpotts and colleagues have not provided that larger study and have not demonstrated that the excision criteria of Liberman et al need to be broadened to include all patients with LCIS.
REFERENCES
Department of Diagnostic Radiology, Yale University School of Medicine, >333 Cedar Street, PO Box 208042, New Haven, CT 06520, e-mail: philpotts@biomed.med.yale.edu
We appreciate Dr Guenin’s interest in our recent article (1). His concern over our recommendation to excise LCIS diagnosed at stereotactic biopsy, however, must be tempered by reviewing the available data and literature.
Diagnosis of LCIS is a rare occurrence at percutaneous core biopsy. In our database of findings from 1,236 stereotactic biopsies, we encountered this entity in only five (0.4%) cases. Liberman et al (2) found LCIS in 16 (1.2%) of 1,315 cases. In a multi-institutional study (3) of findings from 32,424 percutaneous biopsies, LCIS was found in 89 (0.3%).
While the diagnosis of LCIS is a rare occurrence at core biopsy, the rate of malignancy at excision of these lesions is relatively high. Our rate of 20% (one of five lesions) is similar to the rate of 21% by Liberman et al (three of 14 lesions). In the multi-institutional study, malignancy was found in 34% (20 of 58 lesions). While details of the mammographic and histologic features of the cases in the multi-institutional study are not yet available, it is clear that malignancy will be diagnosed at excision in a large proportion of cases.
We fully realize that LCIS is considered an incidental histologic finding at biopsy, and thus surgical excision of these lesions may not make inherent sense. However, there appears to be a high rate of associated high-risk or malignant lesions at percutaneous and surgical biopsy. Four (80%) of five patients in our study had atypical hyperplasia. Nine (64%) of 14 cases in the study by Liberman et al had associated high-risk lesions or features of DCIS. LCIS has been reported (4) to coexist with malignancies, including comedocarcinoma and invasive ductal carcinoma, at surgical biopsy.
Furthermore, at histologic examination, LCIS can have similarities with both atypical lobular hyperplasia and DCIS. Differentiation with core samples may be difficult. In our study, in the one patient with LCIS alone without associated atypical hyperplasia, DCIS was proved with LCIS at excision. Sampling error likely played a role in the underestimation of malignancy in that case, as there were no overlapping features of DCIS. Thus, despite excellent breast pathologic expertise, core biopsy findings may cause malignancy to be underestimated or missed in these cases.
Dr Guenin addresses the issue of complete removal of the mammographic lesion with vacuum-assisted suction biopsy. The underestimation of cancer with the use of this device has been shown (5–7) to be less than that with the 14-gauge–needle and automatic-gun method. However, Dr Guenin can only speculate that lesions containing LCIS removed with the vacuum-assisted suction device do not require excision because, at present and to our knowledge, there are insufficient data in the literature to support that.
While Dr Guenin prefers the criteria of Liberman et al for the excision of LCIS, he must realize that few patients will actually be spared surgery. Their recommendations for excision of LCIS "when the percutaneous biopsy histologic features overlapped with DCIS, when a high-risk lesion was present, or when there was imaging-histologic discordance" would require that the majority of cases (11 of 14 cases) in their study involve surgery. In only three cases was LCIS considered an incidental finding associated with benign tissue that might not require excision. While Dr Guenin claims that our recommendation to excise LCIS is not warranted because it was based on findings in only one patient, Liberman et al’s recommendation not to excise is based on findings in only three. As Liberman and colleagues state, "accrual of these three cases required review of more than 1300 consecutive lesions. ..." Because LCIS is a rare diagnosis at percutaneous biopsy, those patients with LCIS who might be spared excision are exceedingly rare.
Since LCIS is an incidental histologic finding without mammographic features, the issue of imaging-histologic discordance in these cases is somewhat irrelevant. One could argue that the diagnosis of LCIS could always be considered an imaging-histologic discordance. Regardless of the concordance of the mammographic lesion and the histologic diagnoses, the presence of LCIS in a core biopsy specimen requires additional attention.
We do not disagree with the recommendations of Liberman et al, but we caution against not excising these lesions when benign concordant histologic findings are present. Sampling error or pathologic underestimation may not be recognized. Given the rare occurrence of this lesion, the high rate of malignancy at excision, the association with high-risk or malignant lesions, the difficulty in determining imaging-histologic discordance, and the overlapping histologic features with DCIS, we again state that it appears prudent to excise these lesions when they are found at stereotactic biopsy. The burden on the total population of patients would be exceedingly small, yet the implications of missing cancer are substantial.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
