| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ORIGINAL ARTICLES |
From the University of Oklahoma Breast Institute, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma.
Correspondence: Address correspondence and reprint requests to: William C. Dooley, MD, University of Oklahoma Breast Institute, 825 N.E. 10th Street, Suite 3500, University of Oklahoma Health Sciences Center, Oklahoma City, OK 73104; Fax: 405-271-3495; E-mail: wcdooley{at}cox.net
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Methods: All women undergoing lumpectomy in a single surgeon’s practice from January 2000 to August 2001 were evaluated for fluid production from the nipple at the time of lumpectomy. All fluid-producing ducts were cannulated and endoscoped with a 0.9-mm Acueity microendoscope.
Results: Of the 201 patients (16 with atypical ductal hyperplasia, 52 with ductal carcinoma-in-situ, and 133 with stage 1 or 2 breast cancers), 150 (74.6%) could be successfully dilated and scoped. Additional lesions outside the anticipated lumpectomy were identified in 41% (n = 83) of cases. If successful, the chances for a positive margin for cancer decreased from 23.5% to only 5.0%. Endoscopy proved to be a useful adjunct in this series of patients because it identified all cases of extensive intraductal component in early-stage breast cancer.
Conclusions: Routine operative breast endoscopy can reduce the need for re-excision lumpectomy. It also finds substantially more cancerous and precancerous disease than anticipated by routine preoperative mammography and ultrasound.
Key Words: Breast carcinoma • Breast ductoscopy • Breast endoscopy • Breast conservation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Increasingly, the diagnosis has been made through nonsurgical biopsy methods (usually image-guided core biopsy) so that the initial surgery does have therapeutic intent. The ideal would be that the surgeon could be so well guided by pre- and perioperative imaging that second and third attempts to achieve microscopically clear margins would be unnecessary. Unfortunately for most patients, 20% to 35% of the time additional surgery is required after the first lumpectomy attempt because of underestimation of the extent of disease by pre- and perioperative imaging.1,3 The ability to assess margin and disease extent accurately during surgery would considerably advance the research efforts to develop minimally invasive methods of lumpectomy, such as cryotherapy, radiofrequency, or microwave in vivo ablation, in which the loss of traditional pathologic assessment may lead to substantial undertreatment of a small but significant group of patients.
Breast endoscopy was first attempted as an intraoperative tool for the assessment of patients with bloody nipple discharge in the Orient in the early 1990s.4–7 The initial scopes were costly, and the images were poor. Further, the techniques of distending the duct for the relatively large scopes (>1.5-mm outer diameter) were inadequate for exploring more than the initial 1 to 3 cm of the problematic duct. American attempts at breast endoscopy with similar equipment led to the recognition that ductal washings from the scope had considerably more cellular material than prior ductal screening approaches, such as nipple aspiration.8 After the development of a catheter specifically designed to maximize the cytological yield of ductal lavage, the new technique was shown to identify cellular atypia in 24% of high-risk patients.9 Further, a few radiographically occult DCIS lesions were found in patients with the most severe atypia. Techniques were developed to identify each fluid-producing duct in the nonlactating woman’s breast, successfully cannulate those ducts, and distend the duct diffusely to allow for the most complete ductal lavage sampling for this ductal lavage study. With use of these same techniques, most patients with core biopsy–proven atypical ductal hyperplasia (ADH), DCIS, or stage 1 or 2 breast cancer were also found to make fluid for a ductal orifice on the nipple corresponding to a duct connecting to their proliferative breast lesion. The aim of this study was to test the ability of routine operative breast endoscopy to assist the surgeon at intraoperative margin assessment and to define the ductal anatomy in order to best position the lumpectomy to achieve clear margins at first excision of ADH, DCIS, or stage 1 or 2 breast cancer.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Before preparation and draping in the operating room, the nipple was cleansed with a facial abrasive exfoliant. A skin-moisturizing cream was then applied to the breast, and the breast was deeply massaged from the periphery to the center as if expressing milk at lactation. The amount and duration of massage were similar to those commonly used for sentinel node dye migration. The difference was that the massage was performed from periphery to center to fill the lactiferous sinuses with any fluid from the more peripheral ductal tree. This change in the massage seemed to not affect the sentinel node dye migration in any way. Target ducts producing fluid in the operative breast were identified by manual compression of the individual lactiferous sinuses, by use of a standard nipple aspirator, or both. The ductal orifice producing fluid in the same quadrant as the target lesion (and often, in the case of isosulfan blue [LymphazurinTM; Ben Venue Laboratories Inc., Bedford, OH] peritumoral injection for sentinel node, the blue dye producer) was chosen for routine operative breast endoscopy. The duct was cannulated with a 4-O lachrymal duct dilator or a 2-O ProleneTM suture (Ethicon, Somerville, NJ). The duct was then dilated with the Seldinger technique over the indwelling 2-O Prolene by using 26-, 24-, and 22-gauge angiocatheters. When the duct was dilated with the last angiocatheter, then 5 to 10 mL of local anesthetic was injected into the ductal system. After 2 to 3 minutes, an Acueity (Palo Alto, CA) 0.9-mm ductoscope was then advanced into the lactiferous sinus. The ductal system was distended by using saline or further local anesthetic. The duct was then explored, with priority on the largest ductal branches, until every duct larger than the scope had been explored. Rarely, a duct would branch at such an acute angle that the branch orifice could not be cannulated, but this was minimized by constant distraction of the nipple by the operative surgeon to straighten ducts.
All intraluminal lesions seen in the same ductal system as the primary lesion were included within the lumpectomy. To identify lesions, the light was used to simply transilluminate the breast and direct excision. When nonpalpable lesions had been needle-localized before surgery, often the needle transecting the duct or the dye injected to mark the needle tip was visible endoscopically. When specific lesions were identified that had no known pathologic correlate, these were excised separate from the lumpectomy specimen and submitted separately. Once lesions blocked the ducts, further endoscopy was not feasible. All lumpectomy specimens were marked by the author with a six-color Davidson marking system for correlation of pathologic margins with endoscopic findings. Further, this allowed accurate margin assessment in the few cases in which specific endoscopic lesions were removed from the lumpectomy mass separately. The pathologic assessment of margin was made relative to each of these inked margins and determined to be either a nipple-ward or non–nipple-ward margin on the basis of lesion position in the breast.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
In 83 cases (41% of the overall 201), additional intraluminal lesions were seen outside the anticipated surgical margin of 1 cm beyond the lesion on the basis of clinical and preoperative imaging. In the remainder of cases without extra intraluminal disease, margins were minimized by using the scope assistance and that of preoperative needle localization if the lesion was not palpable. In the cases in which additional disease was identified, the axis and extent of lumpectomy were determined on the basis of the plan to remove all endoscopically visible disease. This significantly decreased the use of Langer’s line-based curvilinear incisions in cases in which the additional disease was widespread. In such cases, the use of radial incisions and, rarely, oncoplastic techniques minimized the cosmetic deformity. Even in the case of apparent multiple noncontiguous mammographic lesions, all were found in successful endoscopy cases to be connected to the same ductal system. In at least three cases, the use of endoscopy allowed successful clear margin lumpectomy where the prior radiographical imaging had suggested disease so scattered as to have traditionally required mastectomy.
Margins were positive in 51 patients; these were technical failures of an attempt at endoscopy 23.5% (n = 12) of the time. In the 150 patients undergoing successful endoscopy, the positive-margin rate decreased to 5.0%. The margin positivity in the unsuccessfully scoped cases was evenly distributed in the nipple-ward and non–nipple-ward directions. In the successfully endoscoped patients, the nipple-ward margin was positive in only one case (0.7%), and all other positive margins were in the non–nipple-ward direction. This particular case of positive nipple-ward margin was one of the rare occasions of failure to navigate all the ducts because of acute branch angles. The primary lesion in this case was only 3 to 4 cm from the nipple in the upper outer quadrant, and several large branches came off a very short lactiferous sinus at very acute angles. One of these was not successfully navigated at the time of lumpectomy on the inferolateral position. The margin was positive for a noncontiguous focus of DCIS in this nipple-ward inferolateral position on final pathology of the lumpectomy. All other ductal branches in this region had been endoscopically normal, with smooth white linings.
The frequency of additional intraluminal disease was not significantly different for ADH, DCIS, or invasive ductal carcinoma. There was a trend toward fewer additional intraluminal findings for invasive lobular carcinoma, but because of many fewer cases, this did not reach statistical significance.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The ability to see proliferative disease in the larger breast ducts by endoscopy allows far better determination of extensive intraductal component and ADH extent than that with traditional mammographic and ultrasound imaging.1–3 Faverly et al.16 recently reported that in cases of invasive ductal carcinoma at mastectomy, 53% of patients had localized disease to the region within 1 cm of the edge of the dominant target. This includes some patients classified as having nonlocalized disease because of lymphovascular invasion that extended beyond the classic 1-cm margin. Our findings of 41% of cases with a multifocal invasive, extensive intraductal component and/or regionally associated ADH are in fact very close to their estimation of 47% having disease extending >1 cm from the dominant mass. When less than a mastectomy is performed, we are presuming if margins are clear that the possibility of residual disease is small. In fact, we know that because most breast conservation failures are in the same quadrant as the initial primary lesion, it is also reasonable to speculate that a major portion of these failures are from assuming that all the disease is removed when we have clear margins of a few millimeters. Further, long distances (>1 cm) between individual intraluminal lesions are commonly seen with endoscopy, yet our standard lumpectomy technique ignores the anatomy of the ductal system and the common multifocal nature of the disease within individual ductal trees. Perhaps more complete excision of the proliferative disease associated with a breast cancer could lead to less reliance on radiotherapy as a mandatory adjunct to surgical lumpectomy.
If we are to adopt anatomically directed lumpectomy that removes all of the ductal system associated with peritumoral proliferative disease, further research needs to define the best approach. We did not look at the sum total volume of breast tissue excised relative to the pathologic volume of invasive cancer and DCIS in patients with successful versus unsuccessful endoscopy. To do this and therefore understand the effect of endoscopy on adequate lumpectomy volume will require some standardization of gross and microscopic processing of lumpectomy specimens. Because we found more disease with endoscopy and because the need for the more cosmetically devastating re-excision was less, the cosmesis was subjectively better than with traditional lumpectomy approaches. Further research is needed to judge this more objectively through the patients’ eyes and by volumetric measurement from the digital photography that some plastic surgeons use before breast reconstruction.
The technique of routine operative endoscopy is easily taught and becomes quite practical in most surgeons’ hands after the first 20 to 30 cases.15 Whether or not this technique will affect the rate and pattern of local failure after lumpectomy has yet to be defined. The long-term effects of this change in operative technique will need to be assessed in large multicenter trials. The cannulation of the lactiferous sinus atraumatically enough to allow the local anesthetic to relax and distend the ducts successfully remains a major challenge, especially for the novice. Technical improvements in this most important step are needed before the dedicated high-volume surgeon can embrace routine operative endoscopy. In this respect, it is quite similar to the early sentinel node experience, in which an increasing case number did affect outcome and success with the technique.17
In conclusion, routine operative breast endoscopy can be a major adjunct in breast-conservation management. Decreasing the need for re-excision allows for better cosmetic results. The ability to assess margins in vivo with a reasonable degree of accuracy allows the major problem of minimally invasive ablation of breast cancer to be addressed. Although further study to fully explore its potential is needed, ductoscopy can now become a common and useful tool to the breast surgical oncologist in breast-conservation therapies.
Received for publication March 22, 2002. Accepted for publication August 21, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. S. Dua, C. M. Isacke, and G. P.H. Gui The Intraductal Approach to Breast Cancer Biomarker Discovery J. Clin. Oncol., March 1, 2006; 24(7): 1209 - 1216. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. C. Dooley Ductal Lavage: Risk Stratification Versus Cancer Detection Ann. Surg. Oncol., September 1, 2005; 12(9): 681 - 682. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
