Touch Preparation Cytology of Breast Lumpectomy Margins With Histologic Correlation Charles E. Cox, MD; Ni Ni Ku, MD; Douglas S. Reintgen, MD; Harvey M. Santo V. Nicosia, MD; Stephen Wangensteen, MD
Greenberg, MD;
microscopic disease after lumpectomy for breast significant local recurrence. We evaluated one hundred fourteen consecutive breast lumpectomy margins in this study by touch preparation cytology. Cytologic preparations were intraoperatively correlated with gross and frozen section results and subsequently with permanent histologic sections of representative margins. Three specimens were cytologically unsatisfactory and 86 yielded benign findings, while material suggestive or diagnostic of malignancy was obtained from 25 specimens. Gross, frozen section, and permanent histologic margins were positive in 10, 17, and 22 cases, respectively. There were three false-positive touch preparation cytologic results, while frozen section specimens were false-negative in five cases. Sensitivity and specificity of touch preparation cytology were 100% and 96.6%, respectively, with a diagnostic accuracy of 97.3%. Touch preparation cytologic examination rapidly and reliably evaluates lumpectomy margins and overcomes sampling errors and artifacts related to frozen section evaluation. This technique currently complements frozen section evaluation of lumpectomy margins as part of a protocol aimed at reducing local recurrence
local control.3 The careful evaluation of pathologic margins in breast cancer lumpectomy specimens is fraught with several serious problems. Gross evaluation will lead to an area close to the tumor margin, while frozen section examination provides an accurate assessment of grossly defined margins. However, these methods assess little more than 10% to 15% of the surface area in a given lumpectomy specimen within the time constraints of intraoperative evaluation. Carter4 has proposed parallel sectioning of the entire outer surface of lumpectomy specimens for complete histologie evaluation. Fisher et al5 have recommended sectioning lumpectomy specimens in sev¬ eral different planes to topographically localize tumor cells in representative margins. These techniques, while accurate, analyze surfaces on permanent histologie sections. Gal6 has applied the scraping technique of Shidham et al7 for rapid cytologie evaluation of the resection cavity. This study intro¬ duces touch preparation cytology as a rapid alternative tech¬ nique for evaluating microscopic disease in breast lumpec¬
\s=b\ Residual
cancer may cause
of breast cancer. (Arch Surg. 1991 ;126:490-493)
local rate of up to 25% 6 years after lumpec¬ tomy and radiation therapy may be the result of residual microscopic disease at the time of resection.1,2 Therefore, examination of lumpectomy margins for the presence of tumor cells has become part of many protocols aimed at improving
A
recurrence
Accepted for publication December 31,1990. From the Departments of Surgery (Drs Cox, Reintgen, and Wangensteen), Pathology (Drs Ku and Nicosia), and Radiology (Dr Greenberg), University of South Florida Health Sciences Center, H. Lee Moffitt Cancer Center and Research Institute, Tampa. Read before the 43rd Annual Cancer Symposium of the Society of Surgical
Oncology, Washington, DC, May 20,1990. Reprint requests to the H. Lee Moffitt Cancer Center and Research Institute, University of South Florida, PO Box 280179, Tampa, FL 33682-0179 (Dr Cox).
tomy margins.
MATERIALS AND METHODS We reviewed 162 cases of breast conservation surgery from No¬ vember 1984 through April 1990. From January 1987 to April 1990, 114 patients were evaluated prospectively. The patients' conditions were diagnosed previously using fine-needle aspiration cytology or open biopsies, and included 82 infiltrating ductal carcinomas, 24 intraductal carcinomas, two lobular carcinomas in situ, four infiltrat¬ ing lobular carcinomas, and two mixed infiltrating ductal and lobular carcinomas. The prospective comparison of gross, frozen section, cytologie, and permanent histologie evaluations of breast lumpec¬ tomy margins was done to ascertain diagnostic accuracy of these
techniques.
In the initial phase of our study, we attempted to recover malignant cells from resected margins by dipping the entire specimen in bal¬ anced salt solution. Exfoliated cells were then cytocentrifuged and stained with Papanicolaou's method (Fig 1). However, this technique was not used further because it proved to be time consuming and did not provide an exact topography of involved margins. The protocol ultimately adopted is illustrated in Fig 2.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of British Columbia Library User on 06/19/2015
Lumpectomy
.;
Label Slides
! Rinse in Balanced Salt Solution
Touch Preparation of Margins
Cytocentrifugation
Specimen Orientation I
Air
¡
:
Dry
;
Papanicolaou Stain
Diff-Quik Stain
I
Interpretation I
Intraoperative Report* Fig 1.—Flow chart for cytologie evaluation of lumpectomy margins. The asterisk indicates that the turnaround was less than 15 minutes, and the broken line, the initial phase of the study. Fig 3.—Pie chart of tumor types of 114 specimens.
Table 1.—Correlation Between Gross, Frozen Section, Permanent Histologie, and Touch Preparation Cytologie Examinations of Lumpectomy Margins
Type of Examination
Fig 2.—Schematic diagram of touch preparation cytology of lumpec¬ tomy specimens. Lumpectomy margins were first oriented by various sutures (left) and then sampled by touch preparation cytology (right). S indicates superior; M, medial; D, deep; I, inferior; L, lateral; and T,
Margins True-positive True-negative False-positive False-negative
Gross
Cytologie
Frozen Section
Permanent
10
22
17
22
86
86
89
89
12
tumor.
During surgery, specimen margins were oriented as superior, medial, lateral, inferior, or deep as indicated by various sutures. After slides were labeled, corresponding cellular samples were ob¬ tained from these sites by touch preparation cytology (Fig 2). Ap¬ proximately four to six slides were used for a lumpectomy specimen 5 cm in diameter. An additional touch preparation slide of the tumor itself was used as a positive cytologie control for comparison with the slides of the resected margins. Air-dried samples were stained imme¬ diately using the Diff-Quik method (Baxter Healthcare Corp, Dade Division, Miami, Fia). Final cytologie results were rendered within 15 minutes during surgery. Diagnostic categories used in reporting cytologie findings included unsatisfactory, negative, atypical, suspi¬ cious, or malignant. In the data analysis, atypical cases were included in the negative category, while suspicious cases were considered malignant. Similarly, from a management perspective only, suspi¬ cious and malignant categories were used by the surgeons as an indication to obtain additional margins. RESULTS One hundred fourteen patients ranging in age from 29 to 89 (mean age, 58.36 years) were prospectively evaluated using touch preparation cytology. These included 82 infiltrat¬ ing ductal carcinomas, 24 intraductal carcinomas, two lobular carcinomas in situ, four infiltrating lobular carcinomas, and two mixed infiltrating ductal and lobular carcinomas (Fig 3). Of the 114 lumpectomy specimens evaluated, three were years
categorized as unsatisfactory and excluded from analysis. We noted that the surfaces of the first two specimens were dried due to delay in transportation, while the third specimen ex¬ hibited a folded and retracted surface over an underlying tumor. Tumors extended to gross margins in 10 of 111 cases, while frozen section analysis detected malignancy in 17 cases (Table 1). Touch preparation cytology of lumpectomy margins yield¬ ed 86 benign findings, including 44 with normal ductal cyto¬ logie findings, 25 with fibrocystic changes (eight with associ¬ ated atypical ductal hyperplasia), and 17 with changes of fat necrosis (Figs 4 and 5). There were eight fibrocystic changes with associated atypical ductal hyperplasia. Malignant cellu¬ lar material was identified in 25 lumpectomy margins, includ¬ ing 23 ductal and two lobular carcinomas (Figs 6 through 8).
Examination of permanent histologie specimens revealed positive margins in 22 of 111 lumpectomy specimens. There were three false-positive cases evaluated with touch preparation cytology in which malignant cells were detected in one margin, while frozen and permanent sections of sam¬ pled margins disclosed no tumor involvement. The histologie diagnoses for these cases included one multifocal intraductal carcinoma and two infiltrating ductal carcinomas. In the lat¬ ter case, tumor cells were present within 1 mm of an other¬ wise histologically negative margin. In two cases, malignant cells were correctly identified us¬ ing touch preparation cytologie examination in one margin
Downloaded From: http://archsurg.jamanetwork.com/ by a University of British Columbia Library User on 06/19/2015
Fig 4.—Touch preparation cytologie and histologie features of fibro¬ cystic changes.
Fig 5.—Touch preparation necrosis.
cytologie
and
histologie
Fig 7.—Touch preparation cytologie (left) and corresponding histolog¬ ie (right) features of infiltrating ductal carcinoma (left, Diff-Quik, original magnification 630; right, hematoxylin-eosin, original magnification X100).
features of fat
Fig 8.—Touch preparation cytologie (left) and corresponding histolog¬ ie (right) features of infiltrating lobular carcinoma (left, Diff-Quik, origi¬ nal magnification 400; right, hematoxylin-eosin, original magnifica¬ tion 400).
Table 2—Analysis of Touch Preparation Cytologie Examination of Lumpectomy Margins Variable
Sensitivity Specificity Diagnostic accuracy Positive predictive value Negative predictive value
Fig 6.—Touch preparation cytologie (left) and corresponding histolog¬ (right) features of intraductal carcinoma (left, Diff-Quik, original magnification 630; right, hematoxylin-eosin, original magnification 400). ie
while frozen section and permanent histologie examination showed an infiltrating ductal carcinoma approaching, but not directly involving, the sampled margins. On further review, permanent sections of tumor cells were identified with histo-
% 100
96.6 97.3 88.0 100
logic examination at the cytologically positive margin. Using the predictive value of Galen and Gambino (Table 2), the sensitivity and specificity of touch preparation cytology of lumpectomy margins were 100% and 96.6%, respectively, with an overall diagnostic accuracy of 97.3%. The predictive values for cytologically detected absence and presence of tumor cells were 88.0% and 100%, respectively. COMMENT
study show the following about touch preparation cytology: (1) it topographically evaluates the out¬ er surface of lumpectomy margins; (2) it is highly reliable in The results of this
Downloaded From: http://archsurg.jamanetwork.com/ by a University of British Columbia Library User on 06/19/2015
detecting tumor cells at lumpectomy margins; (3) it rapidly complements and expands the diagnostic role of frozen sec¬ tion evaluation of breast lumpectomy margins; and (4) it reduces the need for repeated resections of histologically involved margins as a separate operative procedure. The need for accurate histologie assessment ofbreast lump¬ ectomy margins in preventing local recurrence has been docu¬ mented. Lagios et al8 reported a local recurrence rate of 28% following lumpectomy without radiation therapy, with 71% of such recurrences associated with positive margins. Harris et al2 reported a local recurrence rate of up to 25% in a retrospec¬ tive study in cases in which margins were positive for intraductal disease. In the National Surgical Adjuvant Breast
Project Prospective Randomized Clinical Trial, Fisher et al5 reported a local recurrence rate of 10% following lumpectomy with radiation therapy when margins were negative. These data suggest the progressive improvement in local recur¬ rence rates with refinement in histologie evaluation of lump¬ ectomy margins. We observe that this further refinement in the evaluation of lumpectomy margins will reduce the local recurrence rate in breast conservation therapy.
Various techniques have been described to evaluate surgi¬ cal margins. Carter's4 method evaluates the entire resection margin in permanent sections. The method of Fisher et al5 is useful in detecting the tumor proximity to margins in repre¬ sentative frozen sections. Shidham et al7 describe a method of scrape cytology that evaluates the entire outer surface of the lumpectomy margin. This last technique may cause cell dam¬ age and harvest cells at some distance from the true margins if vigorous scraping is applied. We attempted to avoid the potential pitfalls of these techniques by applying intraopera¬ tive touch preparation cytology. Using this method, one false-positive case was associated with multifocal intraductal carcinoma, suggesting random harvesting of intraductal foci by cytologie but not histologie evaluation. In the other false-positive cases, it was not clear if the discrepancy between cytologie and histologie findings was due to cells below the true resection margin. When comparing
touch preparation cytology with permanent section histology, it is possible that the touch preparation technique may be more accurate than permanent section evaluation since it evaluates the entire resected surface. In addition, three cases were
categorized as unsatisfactory.
Therefore, touch preparation cytology is not totally free of technical artifacts, such as those related to surface irregular¬ ity and dryness. Subsequent to these artifacts being identi¬ fied, new methods of transport of lumpectomy specimens have been instituted, and this is no longer a problem. How¬ ever, it remains to be seen if such pitfalls may be avoided by other technical modifications, including stretching the lum¬ pectomy surface or aerosol moistening of the surfaces with
balanced salt solution. Tbueh preparations and final cytologie reports of lumpec¬ tomy margins can be rendered within 15 minutes, in contrast to approximately 10 minutes for a single representative fro¬ zen section. Touch preparation cytology is simple and reliable and has high diagnostic accuracy. The technique overcomes frozen section sampling errors and artifacts related to fat and calcifications. The technique further reduces medical costs and the need for reoperation for resection of histologically involved margins. At our institution, touch preparation cytology is an integral part of intraoperative management of breast conservation therapy since we believe this technique identifies microscopic disease not always detectable grossly or by frozen section evaluation. Although not the focus ofthis report, touch prepa¬ ration cytology appears to have affected overall local recur¬ rence rates favorably in this series. Two of 162 patients who underwent lumpectomy have had documented recurrences (mean follow-up, 3.2 years), both of whom had margins evalu¬ ated before the introduction of touch preparation cytology. Prospective studies should confirm our original hypothesis that cytologie evaluation of lumpectomy margins will play an important role in reducing local recurrence with breast con¬ servation therapy.
References 1. Fisher B, Redmond C, Poisson R, Margolese R, et al. Eight-year results of a randomized clinical trial comparing total mastectomy and lumpectomy with or without irradiation in the treatment of breast cancer. N Engl J Med.
1989;320:822-828. 2. Harris JR, Connolly JL, Schnitt SJ, et al. The use of pathologic features in selecting the extent of surgical resection necessary for breast cancer patients by primary radiation therapy. Ann Surg. 1985;201:164-169. 3. Silva EG, Kraemer BB. The examination of margins ofresection by frozen section. Surg Pathol. 1988;1:303-306. 4. Carter D. Margins of 'lumpectomy' for breast cancer. Hum Pathol.
1986;17:330-332. 5. Fisher ER, Sass R, Fisher B, et al. Pathologic findings from the National Surgical Adjuvant Breast Project (Protocol 6), II: relationship of local breast recurrence to multicentricity. Cancer. 1986;57:1717-1724. 6. Gal R. Scrape cytology assessment of margins of lumpectomy specimens in breast cancer. Acta Cytol. 1988;32:838-839. 7. Shidham VB, Dravid NV, Grover S, Kher AV. Role of scrape cytology in rapid intraoperative diagnosis. Acta Cytol. 1976;20:553-555. 8. Lagios MD, Richards VE, Rose MR, Yee E. Segmental mastectomy without radiotherapy. Cancer. 1983;52:2173-2179.
Surgical Anatomy Question
V¿
A
The muscles of facial expression arise from the second branchial arch, are supplied seventh nerve, and are disposed around the apertures of the face. [True or False?]
^tmJ 1. Grant JCB. A Method ofAnatomy. 5th ed.
Baltimore, Md: Williams & Wilkins; 1952:587.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of British Columbia Library User on 06/19/2015
by the
Greenberg, MD;
microscopic disease after lumpectomy for breast significant local recurrence. We evaluated one hundred fourteen consecutive breast lumpectomy margins in this study by touch preparation cytology. Cytologic preparations were intraoperatively correlated with gross and frozen section results and subsequently with permanent histologic sections of representative margins. Three specimens were cytologically unsatisfactory and 86 yielded benign findings, while material suggestive or diagnostic of malignancy was obtained from 25 specimens. Gross, frozen section, and permanent histologic margins were positive in 10, 17, and 22 cases, respectively. There were three false-positive touch preparation cytologic results, while frozen section specimens were false-negative in five cases. Sensitivity and specificity of touch preparation cytology were 100% and 96.6%, respectively, with a diagnostic accuracy of 97.3%. Touch preparation cytologic examination rapidly and reliably evaluates lumpectomy margins and overcomes sampling errors and artifacts related to frozen section evaluation. This technique currently complements frozen section evaluation of lumpectomy margins as part of a protocol aimed at reducing local recurrence
local control.3 The careful evaluation of pathologic margins in breast cancer lumpectomy specimens is fraught with several serious problems. Gross evaluation will lead to an area close to the tumor margin, while frozen section examination provides an accurate assessment of grossly defined margins. However, these methods assess little more than 10% to 15% of the surface area in a given lumpectomy specimen within the time constraints of intraoperative evaluation. Carter4 has proposed parallel sectioning of the entire outer surface of lumpectomy specimens for complete histologie evaluation. Fisher et al5 have recommended sectioning lumpectomy specimens in sev¬ eral different planes to topographically localize tumor cells in representative margins. These techniques, while accurate, analyze surfaces on permanent histologie sections. Gal6 has applied the scraping technique of Shidham et al7 for rapid cytologie evaluation of the resection cavity. This study intro¬ duces touch preparation cytology as a rapid alternative tech¬ nique for evaluating microscopic disease in breast lumpec¬
\s=b\ Residual
cancer may cause
of breast cancer. (Arch Surg. 1991 ;126:490-493)
local rate of up to 25% 6 years after lumpec¬ tomy and radiation therapy may be the result of residual microscopic disease at the time of resection.1,2 Therefore, examination of lumpectomy margins for the presence of tumor cells has become part of many protocols aimed at improving
A
recurrence
Accepted for publication December 31,1990. From the Departments of Surgery (Drs Cox, Reintgen, and Wangensteen), Pathology (Drs Ku and Nicosia), and Radiology (Dr Greenberg), University of South Florida Health Sciences Center, H. Lee Moffitt Cancer Center and Research Institute, Tampa. Read before the 43rd Annual Cancer Symposium of the Society of Surgical
Oncology, Washington, DC, May 20,1990. Reprint requests to the H. Lee Moffitt Cancer Center and Research Institute, University of South Florida, PO Box 280179, Tampa, FL 33682-0179 (Dr Cox).
tomy margins.
MATERIALS AND METHODS We reviewed 162 cases of breast conservation surgery from No¬ vember 1984 through April 1990. From January 1987 to April 1990, 114 patients were evaluated prospectively. The patients' conditions were diagnosed previously using fine-needle aspiration cytology or open biopsies, and included 82 infiltrating ductal carcinomas, 24 intraductal carcinomas, two lobular carcinomas in situ, four infiltrat¬ ing lobular carcinomas, and two mixed infiltrating ductal and lobular carcinomas. The prospective comparison of gross, frozen section, cytologie, and permanent histologie evaluations of breast lumpec¬ tomy margins was done to ascertain diagnostic accuracy of these
techniques.
In the initial phase of our study, we attempted to recover malignant cells from resected margins by dipping the entire specimen in bal¬ anced salt solution. Exfoliated cells were then cytocentrifuged and stained with Papanicolaou's method (Fig 1). However, this technique was not used further because it proved to be time consuming and did not provide an exact topography of involved margins. The protocol ultimately adopted is illustrated in Fig 2.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of British Columbia Library User on 06/19/2015
Lumpectomy
.;
Label Slides
! Rinse in Balanced Salt Solution
Touch Preparation of Margins
Cytocentrifugation
Specimen Orientation I
Air
¡
:
Dry
;
Papanicolaou Stain
Diff-Quik Stain
I
Interpretation I
Intraoperative Report* Fig 1.—Flow chart for cytologie evaluation of lumpectomy margins. The asterisk indicates that the turnaround was less than 15 minutes, and the broken line, the initial phase of the study. Fig 3.—Pie chart of tumor types of 114 specimens.
Table 1.—Correlation Between Gross, Frozen Section, Permanent Histologie, and Touch Preparation Cytologie Examinations of Lumpectomy Margins
Type of Examination
Fig 2.—Schematic diagram of touch preparation cytology of lumpec¬ tomy specimens. Lumpectomy margins were first oriented by various sutures (left) and then sampled by touch preparation cytology (right). S indicates superior; M, medial; D, deep; I, inferior; L, lateral; and T,
Margins True-positive True-negative False-positive False-negative
Gross
Cytologie
Frozen Section
Permanent
10
22
17
22
86
86
89
89
12
tumor.
During surgery, specimen margins were oriented as superior, medial, lateral, inferior, or deep as indicated by various sutures. After slides were labeled, corresponding cellular samples were ob¬ tained from these sites by touch preparation cytology (Fig 2). Ap¬ proximately four to six slides were used for a lumpectomy specimen 5 cm in diameter. An additional touch preparation slide of the tumor itself was used as a positive cytologie control for comparison with the slides of the resected margins. Air-dried samples were stained imme¬ diately using the Diff-Quik method (Baxter Healthcare Corp, Dade Division, Miami, Fia). Final cytologie results were rendered within 15 minutes during surgery. Diagnostic categories used in reporting cytologie findings included unsatisfactory, negative, atypical, suspi¬ cious, or malignant. In the data analysis, atypical cases were included in the negative category, while suspicious cases were considered malignant. Similarly, from a management perspective only, suspi¬ cious and malignant categories were used by the surgeons as an indication to obtain additional margins. RESULTS One hundred fourteen patients ranging in age from 29 to 89 (mean age, 58.36 years) were prospectively evaluated using touch preparation cytology. These included 82 infiltrat¬ ing ductal carcinomas, 24 intraductal carcinomas, two lobular carcinomas in situ, four infiltrating lobular carcinomas, and two mixed infiltrating ductal and lobular carcinomas (Fig 3). Of the 114 lumpectomy specimens evaluated, three were years
categorized as unsatisfactory and excluded from analysis. We noted that the surfaces of the first two specimens were dried due to delay in transportation, while the third specimen ex¬ hibited a folded and retracted surface over an underlying tumor. Tumors extended to gross margins in 10 of 111 cases, while frozen section analysis detected malignancy in 17 cases (Table 1). Touch preparation cytology of lumpectomy margins yield¬ ed 86 benign findings, including 44 with normal ductal cyto¬ logie findings, 25 with fibrocystic changes (eight with associ¬ ated atypical ductal hyperplasia), and 17 with changes of fat necrosis (Figs 4 and 5). There were eight fibrocystic changes with associated atypical ductal hyperplasia. Malignant cellu¬ lar material was identified in 25 lumpectomy margins, includ¬ ing 23 ductal and two lobular carcinomas (Figs 6 through 8).
Examination of permanent histologie specimens revealed positive margins in 22 of 111 lumpectomy specimens. There were three false-positive cases evaluated with touch preparation cytology in which malignant cells were detected in one margin, while frozen and permanent sections of sam¬ pled margins disclosed no tumor involvement. The histologie diagnoses for these cases included one multifocal intraductal carcinoma and two infiltrating ductal carcinomas. In the lat¬ ter case, tumor cells were present within 1 mm of an other¬ wise histologically negative margin. In two cases, malignant cells were correctly identified us¬ ing touch preparation cytologie examination in one margin
Downloaded From: http://archsurg.jamanetwork.com/ by a University of British Columbia Library User on 06/19/2015
Fig 4.—Touch preparation cytologie and histologie features of fibro¬ cystic changes.
Fig 5.—Touch preparation necrosis.
cytologie
and
histologie
Fig 7.—Touch preparation cytologie (left) and corresponding histolog¬ ie (right) features of infiltrating ductal carcinoma (left, Diff-Quik, original magnification 630; right, hematoxylin-eosin, original magnification X100).
features of fat
Fig 8.—Touch preparation cytologie (left) and corresponding histolog¬ ie (right) features of infiltrating lobular carcinoma (left, Diff-Quik, origi¬ nal magnification 400; right, hematoxylin-eosin, original magnifica¬ tion 400).
Table 2—Analysis of Touch Preparation Cytologie Examination of Lumpectomy Margins Variable
Sensitivity Specificity Diagnostic accuracy Positive predictive value Negative predictive value
Fig 6.—Touch preparation cytologie (left) and corresponding histolog¬ (right) features of intraductal carcinoma (left, Diff-Quik, original magnification 630; right, hematoxylin-eosin, original magnification 400). ie
while frozen section and permanent histologie examination showed an infiltrating ductal carcinoma approaching, but not directly involving, the sampled margins. On further review, permanent sections of tumor cells were identified with histo-
% 100
96.6 97.3 88.0 100
logic examination at the cytologically positive margin. Using the predictive value of Galen and Gambino (Table 2), the sensitivity and specificity of touch preparation cytology of lumpectomy margins were 100% and 96.6%, respectively, with an overall diagnostic accuracy of 97.3%. The predictive values for cytologically detected absence and presence of tumor cells were 88.0% and 100%, respectively. COMMENT
study show the following about touch preparation cytology: (1) it topographically evaluates the out¬ er surface of lumpectomy margins; (2) it is highly reliable in The results of this
Downloaded From: http://archsurg.jamanetwork.com/ by a University of British Columbia Library User on 06/19/2015
detecting tumor cells at lumpectomy margins; (3) it rapidly complements and expands the diagnostic role of frozen sec¬ tion evaluation of breast lumpectomy margins; and (4) it reduces the need for repeated resections of histologically involved margins as a separate operative procedure. The need for accurate histologie assessment ofbreast lump¬ ectomy margins in preventing local recurrence has been docu¬ mented. Lagios et al8 reported a local recurrence rate of 28% following lumpectomy without radiation therapy, with 71% of such recurrences associated with positive margins. Harris et al2 reported a local recurrence rate of up to 25% in a retrospec¬ tive study in cases in which margins were positive for intraductal disease. In the National Surgical Adjuvant Breast
Project Prospective Randomized Clinical Trial, Fisher et al5 reported a local recurrence rate of 10% following lumpectomy with radiation therapy when margins were negative. These data suggest the progressive improvement in local recur¬ rence rates with refinement in histologie evaluation of lump¬ ectomy margins. We observe that this further refinement in the evaluation of lumpectomy margins will reduce the local recurrence rate in breast conservation therapy.
Various techniques have been described to evaluate surgi¬ cal margins. Carter's4 method evaluates the entire resection margin in permanent sections. The method of Fisher et al5 is useful in detecting the tumor proximity to margins in repre¬ sentative frozen sections. Shidham et al7 describe a method of scrape cytology that evaluates the entire outer surface of the lumpectomy margin. This last technique may cause cell dam¬ age and harvest cells at some distance from the true margins if vigorous scraping is applied. We attempted to avoid the potential pitfalls of these techniques by applying intraopera¬ tive touch preparation cytology. Using this method, one false-positive case was associated with multifocal intraductal carcinoma, suggesting random harvesting of intraductal foci by cytologie but not histologie evaluation. In the other false-positive cases, it was not clear if the discrepancy between cytologie and histologie findings was due to cells below the true resection margin. When comparing
touch preparation cytology with permanent section histology, it is possible that the touch preparation technique may be more accurate than permanent section evaluation since it evaluates the entire resected surface. In addition, three cases were
categorized as unsatisfactory.
Therefore, touch preparation cytology is not totally free of technical artifacts, such as those related to surface irregular¬ ity and dryness. Subsequent to these artifacts being identi¬ fied, new methods of transport of lumpectomy specimens have been instituted, and this is no longer a problem. How¬ ever, it remains to be seen if such pitfalls may be avoided by other technical modifications, including stretching the lum¬ pectomy surface or aerosol moistening of the surfaces with
balanced salt solution. Tbueh preparations and final cytologie reports of lumpec¬ tomy margins can be rendered within 15 minutes, in contrast to approximately 10 minutes for a single representative fro¬ zen section. Touch preparation cytology is simple and reliable and has high diagnostic accuracy. The technique overcomes frozen section sampling errors and artifacts related to fat and calcifications. The technique further reduces medical costs and the need for reoperation for resection of histologically involved margins. At our institution, touch preparation cytology is an integral part of intraoperative management of breast conservation therapy since we believe this technique identifies microscopic disease not always detectable grossly or by frozen section evaluation. Although not the focus ofthis report, touch prepa¬ ration cytology appears to have affected overall local recur¬ rence rates favorably in this series. Two of 162 patients who underwent lumpectomy have had documented recurrences (mean follow-up, 3.2 years), both of whom had margins evalu¬ ated before the introduction of touch preparation cytology. Prospective studies should confirm our original hypothesis that cytologie evaluation of lumpectomy margins will play an important role in reducing local recurrence with breast con¬ servation therapy.
References 1. Fisher B, Redmond C, Poisson R, Margolese R, et al. Eight-year results of a randomized clinical trial comparing total mastectomy and lumpectomy with or without irradiation in the treatment of breast cancer. N Engl J Med.
1989;320:822-828. 2. Harris JR, Connolly JL, Schnitt SJ, et al. The use of pathologic features in selecting the extent of surgical resection necessary for breast cancer patients by primary radiation therapy. Ann Surg. 1985;201:164-169. 3. Silva EG, Kraemer BB. The examination of margins ofresection by frozen section. Surg Pathol. 1988;1:303-306. 4. Carter D. Margins of 'lumpectomy' for breast cancer. Hum Pathol.
1986;17:330-332. 5. Fisher ER, Sass R, Fisher B, et al. Pathologic findings from the National Surgical Adjuvant Breast Project (Protocol 6), II: relationship of local breast recurrence to multicentricity. Cancer. 1986;57:1717-1724. 6. Gal R. Scrape cytology assessment of margins of lumpectomy specimens in breast cancer. Acta Cytol. 1988;32:838-839. 7. Shidham VB, Dravid NV, Grover S, Kher AV. Role of scrape cytology in rapid intraoperative diagnosis. Acta Cytol. 1976;20:553-555. 8. Lagios MD, Richards VE, Rose MR, Yee E. Segmental mastectomy without radiotherapy. Cancer. 1983;52:2173-2179.
Surgical Anatomy Question
V¿
A
The muscles of facial expression arise from the second branchial arch, are supplied seventh nerve, and are disposed around the apertures of the face. [True or False?]
^tmJ 1. Grant JCB. A Method ofAnatomy. 5th ed.
Baltimore, Md: Williams & Wilkins; 1952:587.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of British Columbia Library User on 06/19/2015
by the
