Practical Radiation Oncology (2013) 3, e107–e112
www.practicalradonc.org
Original Report
Use of postexcision preirradiation mammography in patients with ductal carcinoma in situ of the breast treated with breast-conserving therapy Jonathan T. Whaley MD a , Nataniel H. Lester-Coll MD a , Samantha M. Morrissey a , Abigail B. Milby MD a , Wei-Ting Hwang PhD a , Robert G. Prosnitz MD, MPH b,⁎ a
Department of Radiation Oncology, Hospital of the University of Pennsylvania, Perelman Center for Advanced Medicine, Philadelphia, Pennsylvania b Department of Radiation Oncology, Lehigh Valley Health Network, The John and Dorothy Morgan Cancer Center, Allentown, Pennsylvania Received 30 May 2012; revised 6 August 2012; accepted 17 August 2012
Abstract Purpose: Postexcision preirradiation mammography (PPM) is frequently performed in patients with ductal carcinoma in situ (DCIS) treated with breast-conserving therapy (BCT) to evaluate for residual suspicious calcifications; but no clear evidence supports this practice. The current study was undertaken to investigate the value of PPM in the management of patients with DCIS. Methods and Materials: We conducted a retrospective review of patients treated for DCIS with BCT at the University of Pennsylvania. The impact of PPM on surgical management and on local recurrence was evaluated. Factors associated with the use of PPM, the results of PPM, and the likelihood of finding residual malignancy at the time of re-excision in patients with PPM were also examined. Results: One hundred forty-four of 281 patients (51%) underwent PPM. Of the 144 patients who received PPM, 34 (24%; 95% confidence interval, 17%-31%) had residual suspicious calcifications (a “positive PPM”). Of the 34 patients with a positive PPM, all underwent a re-excision and 19 (56%; 95% confidence interval, 35%-70%) were found to have residual malignancy. Ten of 34 patients with a positive PPM had negative margins, of which 6 had a residual malignancy. Assuming all patients with close, positive, or indeterminate surgical margins would have undergone re-excision regardless of the findings of PPM, PPM resulted in a change in surgical management in 7% (10/144) of patients and removal of residual DCIS in 4% (6/144). With a median follow-up of 9.5 years, the use of PPM was not associated with an improvement in 10-year local recurrence-free survival (94.8% vs 91.5%, P = .368).
This research was presented by Dr Jonathan T. Whaley at the 53rd Annual Meeting of the American Society for Radiation Oncology (ASTRO), Miami Beach, FL, October 2-6, 2011. Conflicts of interest: None. ⁎ Corresponding author. Department of Radiation Oncology, Lehigh Valley Health Network, The John and Dorothy Morgan Cancer Center, 1240 S. Cedar Crest Blvd, Allentown, PA 18103. E-mail address: [email protected] (R.G. Prosnitz). 1879-8500/$ – see front matter © 2013 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.prro.2012.08.002
e108
J.T. Whaley et al
Practical Radiation Oncology: July-September 2013
Conclusions: In this study, PPM affected surgical management in only a small percentage of patients and had no impact on local recurrence. The routine use of PPM in women undergoing BCT for DCIS may not be warranted. © 2013 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.
Introduction Breast-conservation therapy, consisting of wide local excision followed by breast radiation therapy, results in modest rates of local recurrence and excellent cosmesis in patients with ductal carcinoma in situ. 1-6 The success of breast conservation in women relies on adequate surgical excision, such that the burden of residual malignancy is sufficiently small that it can be eliminated by radiation therapy. A microscopically complete excision with negative margins has typically been the criterion used to evaluate the adequacy of breast-conserving surgery (BCS). Table 1 Factors associated with the use of PPM in women with DCIS Variable
PPM (%) (n = 144)
No PPM (%) P value (n = 137)
Age, mean (y) Menopausal status Premenopausal Postmenopausal Perimenopausal Presentation Physical exam (PE) only Mammogram only PE and mammogram Mammographic findings Microcalcifications only Mass only Microcalcifications and mass Other Nuclear grade of DCIS 1-2 3 Margin status Negative Close Positive Histologic subtype Comedo or solid Other Needle localization Yes No Era of treatment Year 2000 and later Before year 2000
57 ± 0.9
59 ± 1.1
.070 .003
27 (19.9) 42 (29.2) 107 78.7) 90 (62.5) 2 (1.4) 12 (8.3) .023 2 (1.4) 6 (4.4) 133 (93.6) 110 (81.5) 7 (5.0) 19 (14.1) b.001 125 (86.8) 79 (57.7) 7 (4.9) 37 (27.0) 9 (6.3) 5 (3.6) 3 (2.1)
16 (11.7)
27 (42.9) 36 (57.1)
27 (34.3) 15 (35.7)
29 (22.8) 60 (47.2) 38 (29.9)
47 (40.2) 46 (39.3) 24 (20.5)
75 (25.1) 69 (47.9)
44 (32.6) 91 (67.4)
0.032
.022
.050
.005 117 (81.3) 91 (66.4) 27 (18.7) 46 (33.6) .629 34 (23.8) 36 (26.3) 109 (76.2) 101 (73.7)
DCIS, ductal carcinoma in situ; PPM, postexcision preirradiation mammography. Values are number (percentage) or mean ± standard error.
Although the adequacy of surgical excision is based on pathologic assessment of the surgical margins, radiographic evaluation also plays an important role in evaluating the completeness of surgical excision. Two types of radiographic evaluation are frequently performed. The first is specimen radiography, in which xrays are obtained of the lumpectomy specimen under compression to ensure the suspicious lesion(s) are contained within the lumpectomy specimen. A second
Table 2 PPM
Univariate analysis of factors associated with a
Variable
Positive PPM Negative PPM P (%) (n = 34) (%) (n = 110) value
Age, mean (y) Menopausal status Premenopausal Postmenopausal Perimenopausal Presentation Physical exam (PE) only Mammogram only PE and mammogram Mammographic findings Microcalcifications only Mass only Microcalcifications and mass Other Nuclear grade of DCIS 1-2 3 Margin status Negative Close Positive Unknown Histologic subtype Comedo or solid Other Needle localization Yes No
57 ± 1.8
57 ± 1.1
11 (32.4) 20 (58.8) 3 (8.8)
31 (28.2) 70 (63.6) 9 (8.2)
0 (0)
2 (1.9)
33 (22.9) 1 (2.9)
100 (92.6) 6 (5.6)
1.000 .876
.641
.804 32 (94.1)
93 (84.5)
1 (2.9) 1 (2.9)
6 (5.5) 8 (7.3)
0 (0)
3 (2.7) .319
10 (47.6) 11 (52.4)
21 (45.7) 25 (54.3)
9 (29.4) 12 (38.2) 9 (26.5) 2 (5.9)
23 (23.7) 47 (48.5) 27 (28.5)
18 (53) 16 (47)
73 (66.4) 37 (33.6)
31(91.2) 3 (8.8)
86 (78.2) 24 (21.8)
.832
.228
.090
DCIS, ductal carcinoma in situ; PPM, postexcision preirradiation mammography. Values are number (percentage) or mean ± standard error.
Practical Radiation Oncology: July-September 2013
PPM in DCIS treated with breast conservation
e109
primary tumor. PPM was performed at the discretion of the treating physician. We did not have information on the use of PPM for 36 patients within our cohort, so the analytic cohort was limited to 281 patients. Due to the
Table 3 Univariate predictors of 10-year local recurrencefree survival (LRFS) in all patients Variable
Figure 1 Kaplan-Meier estimates of local recurrence-free survival in patients managed with and without postexcision preirradiation mammography (PPM). —, without PPM; - - -, with PPM.
measure to ensure the adequate removal of all ductal carcinoma in situ (DCIS) is the assessment for residual suspicious calcifications (RSC) with a postlumpectomy preirradiation mammogram (PPM). Although a number of organizations recommend the routine use of PPM in women undergoing BCT for DCIS, 7 there is no solid evidence that this test confers any significant clinical benefit. Series that have evaluated the clinical utility of PPM in women undergoing breast-conserving therapy (BCT) for DCIS show that changes in management occur in 0%-19% of cases. 8-12 To provide additional data regarding the clinical utility of PPM, we examined the results of PPM in a large cohort of women treated for DCIS with BCT at our institution.
Methods and materials This study was approved by the institutional review board of the University of Pennsylvania. The study cohort consisted of 317 women with DCIS of the breast treated with BCS and breast radiation therapy (RT) at the University of Pennsylvania between 1984 and 2010. Patients were identified from a prospective database of all women treated at our institution with breast BCS and RT for either early-stage invasive breast cancer (IBC) or DCIS from 1978 to the present. The median follow-up time for all 281 women was 9.5 years (range, 0.2-27 years). Of the 317 patients who had undergone BCT for DCIS, information was available on the use of PPM in 281 patients; there were no other exclusion criteria. All patients had histologic confirmation of their disease by the pathology department at our institution, although 40% had their initial surgical procedure performed at outside facilities. The surgical treatment in all cases included complete gross excision of the
Age, y N55 40-55 b40 Menopausal status Postmenopausal Premenopausal Type of presentation Mammographic abnormality Mass or symptoms Mammographic findings Calcifications Calcifications and mass Size of region of calcifications on mammogram, (cm) b1 cm 1-2.5 cm N5 cm DCIS nuclear grade Grade 1-2 Grade 3 Final margin status Negative Close (b2 mm) Positive Comedo necrosis present (%) Absent Present Microinvasion present (%) Absent Present PPM Done Not done Hormonal therapy Received tamoxifen No hormonal therapy Re-excision Done Not done
10-Year LRFS
95% CI
96% 90% 90%
91%-98% 79%-95% 47%-99%
96% 85%
92%-98% 71%-93%
95%
90%-97%
88%
39%-98%
P value .486
.460
.414
.551 94% 92%
88%-97% 79%-97% .650
92% 89% 100%
80%-97% 72%-96%
92% 87%
54%-99% 71%-94%
94% 91% 92%
88%-97% 78%-97% 57-99%
.272
.374
.467 94% 92%
88%-97% 83%-97% .004
94% 56%
90%-97% 73%-88%
95% 91%
89%-98% 82%-96%
97% 93%
82%-99% 87%-96%
94% 93%
63%-99% 88%-96%
.369
.680
.743
CI, confidence interval; DCIS, ductal carcinoma in situ; PPM, postexcision preirradiation mammography. Values are number (percentage) or mean ± standard error.
e110
J.T. Whaley et al
Table 4
Practical Radiation Oncology: July-September 2013
Series evaluating the clinical utility of PPM in women with DCIS
Reference
No.
Years
% DCIS
Margins assessed
Specimen X-ray done
% PPM done
% +PPM
Aref et al 8 Current study Grann et al 10 Kestin et al 13 Waddell et al 12
90 281 101 172 88
1992-97 1984-2010 1999-2001 1980-93 1995-98
44% (40/90) 100% (281/281) 21% (21/101) 100% (172/172) 100% (88/88)
Yes Yes Yes Yes Yes
All patients Some None NR Some
100% (90/90) 51% (144/281) 60% (61/101) 22% (37/172) 76% (67/88)
18% 24% 2% 19% 24%
(16/90) (34/144) (1/61) (7/37) (16/67)
(continued)
retrospective nature of this study, radiologists did not follow a protocol for interpreting PPM; however, standard practice included the acquisition of diagnostic images in standard views and the presence of residual microcalcifications in the quadrant of interest or a suspicious mass other than expected seroma are hereafter referred to as a “positive PPM.” The final pathology margin status was determined retrospectively from patient records. Pathologic margins were considered negative (≥2 mm), close (b2 mm), or positive (tumor at inked margin). Our database did not routinely include information on the use of specimen radiography in this cohort of patients or the results of specimen radiography. The total radiation dose was defined as the sum of the dose from the tangential fields and the dose from the breast boost. Whole breast RT with tangential fields was delivered for all women to a median dose of 50 Gy (range, 42-50) within 4.5-5 weeks. The radiation energy was generally 6-MV photons, but higher energies were used as indicated. During the study period, a boost to the tumor bed was routinely administered. A tumor bed boost was omitted at the discretion of the treating physician in only a single patient who was felt to have low-risk DCIS based on clinical and pathologic criteria. The boost was generally delivered using electrons of varying energy. The median total dose was 63 Gy (range, 50-63 Gy). Adjuvant endocrine therapy with tamoxifen was given in 35 women. Descriptive statistics were produced for all study variables. For the univariate analyses, factors associated with the use of PPM, a positive PPM and a positive reexcision were examined by the Pearson χ 2 test for categoric variables, or the 2-sample t test for continuous variables. The Kaplan-Meier method was used to estimate local recurrence free survival (LRFS), and the log-rank test was used to evaluate candidate predictors of LRFS. A local recurrence (LR) was defined as treatment failure within the treated breast. Time to LR was calculated based on the interval from the start of RT to the development of LR. Patients were censored from the analysis of LRFS if they developed regional recurrence, distant recurrence, or if they died of other causes. A P value b.05 was considered significant. Data were analyzed using STATA, Version 11 (StatCorp, College Station, TX).
Results Predictors of the use of PPM The patient and tumor characteristics, stratified according to the use of PPM, are shown in Table 1. One hundred fortyfour of 281 patients (56%) underwent PPM. Premenopausal status (P = .003), mammographic presentation (P = .023), the use of needle-guided biopsy (P = .005), high nuclear grade (0.03), close or positive margins (P = .003), and comedo subtype (P = .001) were all significantly associated with the use of PPM. In addition, there was a trend toward higher use of PPM with younger age (P = .07).
Predictors of a positive PPM among patients who underwent PPM Of the 144 patients who underwent PPM, 34 (23.6%) were found to have positive PPM. As shown in Table 2, no significant associations between the result of the PPM and any of the variables examined, including size of DCIS, nuclear grade, and margin status were observed on univariate analysis. There was a trend toward higher rates of positive PPM among patients who had their initial surgery conducted with the aid of needle localization versus those who had undergone excision of palpable lesions (78.2% vs 21.8%, P = .09). Of the 34 patients with a positive PPM, specimen radiography had been performed in 26 (76%); 24 (92%) of these radiographs documented excision of the microcalcifications of interest and only 2 (8%) suggested incomplete removal of microcalcifications. Among the 24 patients who had specimen radiographs indicating complete excision of the microcalcifications of interest, 17 had close or positive margins and 7 had negative margins. Of the 2 patients who had specimen radiographs that suggested incomplete removal of the microcalcifications of interest, both had positive margins.
Impact of PPM on local recurrence With a median follow-up of 9.5 years, the 10-year local LRFS for all 281 patients in the cohort was 93.4%. The median time to LR was 7.1 years. There was no difference in the 10-year rates of LRFS among patients who did and did not receive a PPM (94.8% vs 91.5%, P = .368). This lack of
Practical Radiation Oncology: July-September 2013
PPM in DCIS treated with breast conservation
e111
Table 4 (continued) Reference
% Pathologic % Change validation management
% Residual malignancy % Residual malignancy % Residual malignancy Impact (+PPM) (all PPM) found due to PPM on LR
75% (12/16) 10% (9/90) 67% (8/12) Aref et al 8 Current study 100% (34/34) 7% (10/144) 56% (19/34) Grann et al 10 100% (1/1) 2% (1/61) 100% (1/1) Kestin et al 13 100% (7/7) 19% (7/37) 86% (6/7) Waddell et al 12 88% (14/16) 9% (6/67) 64% (9/14)
9% (8/90) 13% (19/144) 2% (1/61) 16% (6/37) 13% (9/67)
9% (8/90) 4% (6/144) 2% (1/61) NR 4% (3/67)
NR No No No NR
DCIS, ductal carcinoma in situ; LR, local recurrence; NR, not reported; PPM, preirradiation mammography.
difference in LRFS was expected with such a high 10-year LRFS (91.5%) in the control arm. Kaplan-Meier estimates of LRFS for patients who did and did not receive a postexcision preirradiation mammogram are shown in Fig 1. In sum, 20 patients, 10 in each group, had local or regional recurrence during the follow-up period. Of the 2 patients with positive PPM and local recurrence, 1 patient had re-excision for residual DCIS and the other was biopsy negative despite residuals calcifications. On univariate analysis, only microinvasive carcinoma was associated with a lower 10-year LRFS (94% vs 56%, P = .004). These results are shown in Table 3.
Impact of PPM on surgical management and surgical outcomes Of the 34 patients who had a positive PPM, only 10 of these patients had negative margins. Assuming the other 24 patients with positive, close, or unknown margin status would have undergone re-excision regardless of the PPM findings, additional surgical management was undertaken in only 7% (10/144) of patients who underwent PPM. Of these 10 patients, 6 (4% of patients who underwent PPM) had residual malignancy.
Discussion The clinical utility of PPM in women with DCIS undergoing BCT is unclear. Although the American College of Radiology, American College of Surgeons, College of American Pathology, and Society of Surgical Oncology recommend that women with DCIS undergo a PPM to document the removal of residual calcifications unless specimen radiograph clearly documents removal of all calcifications, 7 solid evidence confirming the necessity of obtaining a PPM in women with DCIS is lacking. As surgical and pathologic techniques have improved and now routinely include the use of needle localization and specimen radiography, many physicians choose to forego PPM and rely on the pathologic specimen margins for treatment decisions. However, in light of the increasing use of partial breast irradiation and of the omission of RT in elderly women following lumpectomy for small
estrogen-receptor positive DCIS, the utility of PPM is an increasingly important issue. In the current study, 51% (144/281) patients treated with BCT for DCIS at the University of Pennsylvania from 1984 to 2010 underwent a PPM. The clinicopathologic features of patients who did and did not undergo PPM differed significantly, which suggests that selection bias played a significant role in determining which patients underwent this evaluation. Thirty-four of 144 (24%) patients who received a PPM were found to have RSC, but only 10 of these patients had negative initial surgical margins. Assuming that all patients with close, positive, or unknown margins would have undergone re-excision regardless of the findings of a PPM, the PPM led to a change in surgical management in only 7% of patients in whom this test was performed and a finding of residual malignancy in 4%. Four previous publications have reported the results of PPM in women treated for either IBC or DCIS with BCT (see Table 4). 8,10,12,13 In these studies the proportion of patients found to have a positive PPM ranged from 2%24%. PPM led to a change in surgical management in 2%19% of patients. However, only 2%-9% of patients were found to have residual malignancy on re-excision performed exclusively because of the PPM findings (ie, there was no other rationale for re-excision such as close or positive margins). Although the findings of the current study are similar to those of these prior publications, the current study had a much larger sample size and was restricted to patients with DCIS. Our study has several limitations. First, our results are subject to selection bias. As PPMs were not routinely performed in all patients, it is likely that PPMs were preferentially obtained in patients with a higher likelihood of having residual disease in the breast after initial excision (eg, patients with close or positive margins). Although we did not observe an improvement in LRFS in patients undergoing PPM, it is possible the imbalance of adverse prognostic factors in the PPM group obscured what would have been a statistically significant improvement in LRFS in this group. Second, our database lacked complete information on some variables, which may have influenced the likelihood of obtaining a positive PPM such as the extent of calcifications on the initial mammogram, the use of specimen radiography, and the findings of specimen radiography. Third, the test characteristics of PPM
e112
J.T. Whaley et al
(sensitivity, specificity, positive predictive value, and negative predictive value) cannot be calculated from our data set as re-excisions were not performed in all patients with negative PPMs.
Practical Radiation Oncology: July-September 2013
2.
3.
Conclusions PPM appears to result in a change in surgical management in a small percentage of patients treated for DCIS with BCT and excision of additional malignancy in an even smaller proportion. Furthermore, PPM does not appear to have any measurable impact on LRFS. The reasons underlying the apparently modest impact of PPM are clear. When the surgical margins are close or positive or the specimen radiograph does not clearly demonstrate the suspicious calcifications seen on the screening or diagnostic mammograms, re-excision is likely to be performed regardless of the results of a PPM. Conversely, when the surgical margins are negative and the specimen radiograph identifies the calcifications of interest, PPM is unlikely to be positive. Furthermore, in the small subset of these patients who would be found to have RSC on a PPM, whole breast radiation therapy may well be sufficient to sterilize any minute amounts of residual disease associated with these microcalcifications. It is well established, for example, that the proportion of patients found to have multifocal disease on breast magnetic resonance imaging (16%) far exceeds the observed local recurrence rate following BCT. 14 In institutions where careful assessment of pathologic margins is performed and specimen radiography is routinely performed, the current study suggests that the routine PPM is not warranted.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
References 1. Bijker N, Meijnen P, Peterse JL, et al. Breast-conserving treatment with or without radiotherapy in ductal carcinoma-in-situ: ten-year results of European Organisation for Research and Treatment of Cancer randomized phase III trial 10853–a study by the EORTC
14.
Breast Cancer Cooperative Group and EORTC Radiotherapy Group. J Clin Oncol. 2006;24:3381-3387. Fisher B, Dignam J, Wolmark N, et al. Lumpectomy and radiation therapy for the treatment of intraductal breast cancer: findings from National Surgical Adjuvant Breast and Bowel Project B-17. J Clin Oncol. 1998;16:441-452. Holmberg L, Garmo H, Granstrand B, et al. Absolute risk reductions for local recurrence after postoperative radiotherapy after sector resection for ductal carcinoma in situ of the breast. J Clin Oncol. 2008;26:1247-1252. Macdonald HR, Silverstein MJ, Lee LA, et al. Margin width as the sole determinant of local recurrence after breast conservation in patients with ductal carcinoma in situ of the breast. Am J Surg. 2006;192:420-422. Solin LJ, Fourquet A, Vicini FA, et al. Mammographically detected ductal carcinoma in situ of the breast treated with breast-conserving surgery and definitive breast irradiation: long-term outcome and prognostic significance of patient age and margin status. Int J Radiat Oncol Biol Phys. 2001;50:991-1002. Solin LJ, Fourquet A, Vicini FA, et al. Long-term outcome after breastconservation treatment with radiation for mammographically detected ductal carcinoma in situ of the breast. Cancer. 2005;103:1137-1146. Morrow M, Strom EA, Bassett LW, et al. Standard for the management of ductal carcinoma in situ of the breast (DCIS). CA Cancer J Clin. 2002;52:256-276. Aref A, Youssef E, Washington T, et al. The value of postlumpectomy mammogram in the management of breast cancer patients presenting with suspiciouis microcalcifications. Cancer J Sci Am. 2000;6:25-27. Gluck BS, Dershaw DD, Liberman L, Deutch BM. Microcalcifications on postoperative mammograms as an indicator of adequacy of tumor excision. Radiology. 1993;188:469-472. Grann A, Abdou JC, Dragman N, Goodman R. The value of postexcision preradiation mammography in patients with early-stage breast cancer. Am J Clin Oncol. 2004;27:285-288. Teixidor HS, Chu FC, Kim YS, Levin TL. The value of mammography after limited breast surgery and before definitive radiation therapy. Cancer. 1992;69:1418-1423. Waddell BE, Stomper PC, DeFazio JL, Hurd TC, Edge SB. Postexcision mammography is indicated after resection of ductal carcinoma-in-situ of the breast. Ann Surg Oncol. 2000;7: 665-668. Kestin LL, Goldstein NS, Martinez AA, et al. Mammographically detected ductal carcinoma in situ treated with conservative surgery with or without radiation therapy: patterns of failure and 10-year results. Ann Surg. 2000;231:235-245. Houssami N, Ciatto S, Macaskill P, et al. Accuracy and surgical impact of magnetic resonance imaging in breast cancer staging: systematic review and meta-analysis in detection of multifocal and multicentric cancer. J Clin Oncol. 2008;26:3248-3258.
www.practicalradonc.org
Original Report
Use of postexcision preirradiation mammography in patients with ductal carcinoma in situ of the breast treated with breast-conserving therapy Jonathan T. Whaley MD a , Nataniel H. Lester-Coll MD a , Samantha M. Morrissey a , Abigail B. Milby MD a , Wei-Ting Hwang PhD a , Robert G. Prosnitz MD, MPH b,⁎ a
Department of Radiation Oncology, Hospital of the University of Pennsylvania, Perelman Center for Advanced Medicine, Philadelphia, Pennsylvania b Department of Radiation Oncology, Lehigh Valley Health Network, The John and Dorothy Morgan Cancer Center, Allentown, Pennsylvania Received 30 May 2012; revised 6 August 2012; accepted 17 August 2012
Abstract Purpose: Postexcision preirradiation mammography (PPM) is frequently performed in patients with ductal carcinoma in situ (DCIS) treated with breast-conserving therapy (BCT) to evaluate for residual suspicious calcifications; but no clear evidence supports this practice. The current study was undertaken to investigate the value of PPM in the management of patients with DCIS. Methods and Materials: We conducted a retrospective review of patients treated for DCIS with BCT at the University of Pennsylvania. The impact of PPM on surgical management and on local recurrence was evaluated. Factors associated with the use of PPM, the results of PPM, and the likelihood of finding residual malignancy at the time of re-excision in patients with PPM were also examined. Results: One hundred forty-four of 281 patients (51%) underwent PPM. Of the 144 patients who received PPM, 34 (24%; 95% confidence interval, 17%-31%) had residual suspicious calcifications (a “positive PPM”). Of the 34 patients with a positive PPM, all underwent a re-excision and 19 (56%; 95% confidence interval, 35%-70%) were found to have residual malignancy. Ten of 34 patients with a positive PPM had negative margins, of which 6 had a residual malignancy. Assuming all patients with close, positive, or indeterminate surgical margins would have undergone re-excision regardless of the findings of PPM, PPM resulted in a change in surgical management in 7% (10/144) of patients and removal of residual DCIS in 4% (6/144). With a median follow-up of 9.5 years, the use of PPM was not associated with an improvement in 10-year local recurrence-free survival (94.8% vs 91.5%, P = .368).
This research was presented by Dr Jonathan T. Whaley at the 53rd Annual Meeting of the American Society for Radiation Oncology (ASTRO), Miami Beach, FL, October 2-6, 2011. Conflicts of interest: None. ⁎ Corresponding author. Department of Radiation Oncology, Lehigh Valley Health Network, The John and Dorothy Morgan Cancer Center, 1240 S. Cedar Crest Blvd, Allentown, PA 18103. E-mail address: [email protected] (R.G. Prosnitz). 1879-8500/$ – see front matter © 2013 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.prro.2012.08.002
e108
J.T. Whaley et al
Practical Radiation Oncology: July-September 2013
Conclusions: In this study, PPM affected surgical management in only a small percentage of patients and had no impact on local recurrence. The routine use of PPM in women undergoing BCT for DCIS may not be warranted. © 2013 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.
Introduction Breast-conservation therapy, consisting of wide local excision followed by breast radiation therapy, results in modest rates of local recurrence and excellent cosmesis in patients with ductal carcinoma in situ. 1-6 The success of breast conservation in women relies on adequate surgical excision, such that the burden of residual malignancy is sufficiently small that it can be eliminated by radiation therapy. A microscopically complete excision with negative margins has typically been the criterion used to evaluate the adequacy of breast-conserving surgery (BCS). Table 1 Factors associated with the use of PPM in women with DCIS Variable
PPM (%) (n = 144)
No PPM (%) P value (n = 137)
Age, mean (y) Menopausal status Premenopausal Postmenopausal Perimenopausal Presentation Physical exam (PE) only Mammogram only PE and mammogram Mammographic findings Microcalcifications only Mass only Microcalcifications and mass Other Nuclear grade of DCIS 1-2 3 Margin status Negative Close Positive Histologic subtype Comedo or solid Other Needle localization Yes No Era of treatment Year 2000 and later Before year 2000
57 ± 0.9
59 ± 1.1
.070 .003
27 (19.9) 42 (29.2) 107 78.7) 90 (62.5) 2 (1.4) 12 (8.3) .023 2 (1.4) 6 (4.4) 133 (93.6) 110 (81.5) 7 (5.0) 19 (14.1) b.001 125 (86.8) 79 (57.7) 7 (4.9) 37 (27.0) 9 (6.3) 5 (3.6) 3 (2.1)
16 (11.7)
27 (42.9) 36 (57.1)
27 (34.3) 15 (35.7)
29 (22.8) 60 (47.2) 38 (29.9)
47 (40.2) 46 (39.3) 24 (20.5)
75 (25.1) 69 (47.9)
44 (32.6) 91 (67.4)
0.032
.022
.050
.005 117 (81.3) 91 (66.4) 27 (18.7) 46 (33.6) .629 34 (23.8) 36 (26.3) 109 (76.2) 101 (73.7)
DCIS, ductal carcinoma in situ; PPM, postexcision preirradiation mammography. Values are number (percentage) or mean ± standard error.
Although the adequacy of surgical excision is based on pathologic assessment of the surgical margins, radiographic evaluation also plays an important role in evaluating the completeness of surgical excision. Two types of radiographic evaluation are frequently performed. The first is specimen radiography, in which xrays are obtained of the lumpectomy specimen under compression to ensure the suspicious lesion(s) are contained within the lumpectomy specimen. A second
Table 2 PPM
Univariate analysis of factors associated with a
Variable
Positive PPM Negative PPM P (%) (n = 34) (%) (n = 110) value
Age, mean (y) Menopausal status Premenopausal Postmenopausal Perimenopausal Presentation Physical exam (PE) only Mammogram only PE and mammogram Mammographic findings Microcalcifications only Mass only Microcalcifications and mass Other Nuclear grade of DCIS 1-2 3 Margin status Negative Close Positive Unknown Histologic subtype Comedo or solid Other Needle localization Yes No
57 ± 1.8
57 ± 1.1
11 (32.4) 20 (58.8) 3 (8.8)
31 (28.2) 70 (63.6) 9 (8.2)
0 (0)
2 (1.9)
33 (22.9) 1 (2.9)
100 (92.6) 6 (5.6)
1.000 .876
.641
.804 32 (94.1)
93 (84.5)
1 (2.9) 1 (2.9)
6 (5.5) 8 (7.3)
0 (0)
3 (2.7) .319
10 (47.6) 11 (52.4)
21 (45.7) 25 (54.3)
9 (29.4) 12 (38.2) 9 (26.5) 2 (5.9)
23 (23.7) 47 (48.5) 27 (28.5)
18 (53) 16 (47)
73 (66.4) 37 (33.6)
31(91.2) 3 (8.8)
86 (78.2) 24 (21.8)
.832
.228
.090
DCIS, ductal carcinoma in situ; PPM, postexcision preirradiation mammography. Values are number (percentage) or mean ± standard error.
Practical Radiation Oncology: July-September 2013
PPM in DCIS treated with breast conservation
e109
primary tumor. PPM was performed at the discretion of the treating physician. We did not have information on the use of PPM for 36 patients within our cohort, so the analytic cohort was limited to 281 patients. Due to the
Table 3 Univariate predictors of 10-year local recurrencefree survival (LRFS) in all patients Variable
Figure 1 Kaplan-Meier estimates of local recurrence-free survival in patients managed with and without postexcision preirradiation mammography (PPM). —, without PPM; - - -, with PPM.
measure to ensure the adequate removal of all ductal carcinoma in situ (DCIS) is the assessment for residual suspicious calcifications (RSC) with a postlumpectomy preirradiation mammogram (PPM). Although a number of organizations recommend the routine use of PPM in women undergoing BCT for DCIS, 7 there is no solid evidence that this test confers any significant clinical benefit. Series that have evaluated the clinical utility of PPM in women undergoing breast-conserving therapy (BCT) for DCIS show that changes in management occur in 0%-19% of cases. 8-12 To provide additional data regarding the clinical utility of PPM, we examined the results of PPM in a large cohort of women treated for DCIS with BCT at our institution.
Methods and materials This study was approved by the institutional review board of the University of Pennsylvania. The study cohort consisted of 317 women with DCIS of the breast treated with BCS and breast radiation therapy (RT) at the University of Pennsylvania between 1984 and 2010. Patients were identified from a prospective database of all women treated at our institution with breast BCS and RT for either early-stage invasive breast cancer (IBC) or DCIS from 1978 to the present. The median follow-up time for all 281 women was 9.5 years (range, 0.2-27 years). Of the 317 patients who had undergone BCT for DCIS, information was available on the use of PPM in 281 patients; there were no other exclusion criteria. All patients had histologic confirmation of their disease by the pathology department at our institution, although 40% had their initial surgical procedure performed at outside facilities. The surgical treatment in all cases included complete gross excision of the
Age, y N55 40-55 b40 Menopausal status Postmenopausal Premenopausal Type of presentation Mammographic abnormality Mass or symptoms Mammographic findings Calcifications Calcifications and mass Size of region of calcifications on mammogram, (cm) b1 cm 1-2.5 cm N5 cm DCIS nuclear grade Grade 1-2 Grade 3 Final margin status Negative Close (b2 mm) Positive Comedo necrosis present (%) Absent Present Microinvasion present (%) Absent Present PPM Done Not done Hormonal therapy Received tamoxifen No hormonal therapy Re-excision Done Not done
10-Year LRFS
95% CI
96% 90% 90%
91%-98% 79%-95% 47%-99%
96% 85%
92%-98% 71%-93%
95%
90%-97%
88%
39%-98%
P value .486
.460
.414
.551 94% 92%
88%-97% 79%-97% .650
92% 89% 100%
80%-97% 72%-96%
92% 87%
54%-99% 71%-94%
94% 91% 92%
88%-97% 78%-97% 57-99%
.272
.374
.467 94% 92%
88%-97% 83%-97% .004
94% 56%
90%-97% 73%-88%
95% 91%
89%-98% 82%-96%
97% 93%
82%-99% 87%-96%
94% 93%
63%-99% 88%-96%
.369
.680
.743
CI, confidence interval; DCIS, ductal carcinoma in situ; PPM, postexcision preirradiation mammography. Values are number (percentage) or mean ± standard error.
e110
J.T. Whaley et al
Table 4
Practical Radiation Oncology: July-September 2013
Series evaluating the clinical utility of PPM in women with DCIS
Reference
No.
Years
% DCIS
Margins assessed
Specimen X-ray done
% PPM done
% +PPM
Aref et al 8 Current study Grann et al 10 Kestin et al 13 Waddell et al 12
90 281 101 172 88
1992-97 1984-2010 1999-2001 1980-93 1995-98
44% (40/90) 100% (281/281) 21% (21/101) 100% (172/172) 100% (88/88)
Yes Yes Yes Yes Yes
All patients Some None NR Some
100% (90/90) 51% (144/281) 60% (61/101) 22% (37/172) 76% (67/88)
18% 24% 2% 19% 24%
(16/90) (34/144) (1/61) (7/37) (16/67)
(continued)
retrospective nature of this study, radiologists did not follow a protocol for interpreting PPM; however, standard practice included the acquisition of diagnostic images in standard views and the presence of residual microcalcifications in the quadrant of interest or a suspicious mass other than expected seroma are hereafter referred to as a “positive PPM.” The final pathology margin status was determined retrospectively from patient records. Pathologic margins were considered negative (≥2 mm), close (b2 mm), or positive (tumor at inked margin). Our database did not routinely include information on the use of specimen radiography in this cohort of patients or the results of specimen radiography. The total radiation dose was defined as the sum of the dose from the tangential fields and the dose from the breast boost. Whole breast RT with tangential fields was delivered for all women to a median dose of 50 Gy (range, 42-50) within 4.5-5 weeks. The radiation energy was generally 6-MV photons, but higher energies were used as indicated. During the study period, a boost to the tumor bed was routinely administered. A tumor bed boost was omitted at the discretion of the treating physician in only a single patient who was felt to have low-risk DCIS based on clinical and pathologic criteria. The boost was generally delivered using electrons of varying energy. The median total dose was 63 Gy (range, 50-63 Gy). Adjuvant endocrine therapy with tamoxifen was given in 35 women. Descriptive statistics were produced for all study variables. For the univariate analyses, factors associated with the use of PPM, a positive PPM and a positive reexcision were examined by the Pearson χ 2 test for categoric variables, or the 2-sample t test for continuous variables. The Kaplan-Meier method was used to estimate local recurrence free survival (LRFS), and the log-rank test was used to evaluate candidate predictors of LRFS. A local recurrence (LR) was defined as treatment failure within the treated breast. Time to LR was calculated based on the interval from the start of RT to the development of LR. Patients were censored from the analysis of LRFS if they developed regional recurrence, distant recurrence, or if they died of other causes. A P value b.05 was considered significant. Data were analyzed using STATA, Version 11 (StatCorp, College Station, TX).
Results Predictors of the use of PPM The patient and tumor characteristics, stratified according to the use of PPM, are shown in Table 1. One hundred fortyfour of 281 patients (56%) underwent PPM. Premenopausal status (P = .003), mammographic presentation (P = .023), the use of needle-guided biopsy (P = .005), high nuclear grade (0.03), close or positive margins (P = .003), and comedo subtype (P = .001) were all significantly associated with the use of PPM. In addition, there was a trend toward higher use of PPM with younger age (P = .07).
Predictors of a positive PPM among patients who underwent PPM Of the 144 patients who underwent PPM, 34 (23.6%) were found to have positive PPM. As shown in Table 2, no significant associations between the result of the PPM and any of the variables examined, including size of DCIS, nuclear grade, and margin status were observed on univariate analysis. There was a trend toward higher rates of positive PPM among patients who had their initial surgery conducted with the aid of needle localization versus those who had undergone excision of palpable lesions (78.2% vs 21.8%, P = .09). Of the 34 patients with a positive PPM, specimen radiography had been performed in 26 (76%); 24 (92%) of these radiographs documented excision of the microcalcifications of interest and only 2 (8%) suggested incomplete removal of microcalcifications. Among the 24 patients who had specimen radiographs indicating complete excision of the microcalcifications of interest, 17 had close or positive margins and 7 had negative margins. Of the 2 patients who had specimen radiographs that suggested incomplete removal of the microcalcifications of interest, both had positive margins.
Impact of PPM on local recurrence With a median follow-up of 9.5 years, the 10-year local LRFS for all 281 patients in the cohort was 93.4%. The median time to LR was 7.1 years. There was no difference in the 10-year rates of LRFS among patients who did and did not receive a PPM (94.8% vs 91.5%, P = .368). This lack of
Practical Radiation Oncology: July-September 2013
PPM in DCIS treated with breast conservation
e111
Table 4 (continued) Reference
% Pathologic % Change validation management
% Residual malignancy % Residual malignancy % Residual malignancy Impact (+PPM) (all PPM) found due to PPM on LR
75% (12/16) 10% (9/90) 67% (8/12) Aref et al 8 Current study 100% (34/34) 7% (10/144) 56% (19/34) Grann et al 10 100% (1/1) 2% (1/61) 100% (1/1) Kestin et al 13 100% (7/7) 19% (7/37) 86% (6/7) Waddell et al 12 88% (14/16) 9% (6/67) 64% (9/14)
9% (8/90) 13% (19/144) 2% (1/61) 16% (6/37) 13% (9/67)
9% (8/90) 4% (6/144) 2% (1/61) NR 4% (3/67)
NR No No No NR
DCIS, ductal carcinoma in situ; LR, local recurrence; NR, not reported; PPM, preirradiation mammography.
difference in LRFS was expected with such a high 10-year LRFS (91.5%) in the control arm. Kaplan-Meier estimates of LRFS for patients who did and did not receive a postexcision preirradiation mammogram are shown in Fig 1. In sum, 20 patients, 10 in each group, had local or regional recurrence during the follow-up period. Of the 2 patients with positive PPM and local recurrence, 1 patient had re-excision for residual DCIS and the other was biopsy negative despite residuals calcifications. On univariate analysis, only microinvasive carcinoma was associated with a lower 10-year LRFS (94% vs 56%, P = .004). These results are shown in Table 3.
Impact of PPM on surgical management and surgical outcomes Of the 34 patients who had a positive PPM, only 10 of these patients had negative margins. Assuming the other 24 patients with positive, close, or unknown margin status would have undergone re-excision regardless of the PPM findings, additional surgical management was undertaken in only 7% (10/144) of patients who underwent PPM. Of these 10 patients, 6 (4% of patients who underwent PPM) had residual malignancy.
Discussion The clinical utility of PPM in women with DCIS undergoing BCT is unclear. Although the American College of Radiology, American College of Surgeons, College of American Pathology, and Society of Surgical Oncology recommend that women with DCIS undergo a PPM to document the removal of residual calcifications unless specimen radiograph clearly documents removal of all calcifications, 7 solid evidence confirming the necessity of obtaining a PPM in women with DCIS is lacking. As surgical and pathologic techniques have improved and now routinely include the use of needle localization and specimen radiography, many physicians choose to forego PPM and rely on the pathologic specimen margins for treatment decisions. However, in light of the increasing use of partial breast irradiation and of the omission of RT in elderly women following lumpectomy for small
estrogen-receptor positive DCIS, the utility of PPM is an increasingly important issue. In the current study, 51% (144/281) patients treated with BCT for DCIS at the University of Pennsylvania from 1984 to 2010 underwent a PPM. The clinicopathologic features of patients who did and did not undergo PPM differed significantly, which suggests that selection bias played a significant role in determining which patients underwent this evaluation. Thirty-four of 144 (24%) patients who received a PPM were found to have RSC, but only 10 of these patients had negative initial surgical margins. Assuming that all patients with close, positive, or unknown margins would have undergone re-excision regardless of the findings of a PPM, the PPM led to a change in surgical management in only 7% of patients in whom this test was performed and a finding of residual malignancy in 4%. Four previous publications have reported the results of PPM in women treated for either IBC or DCIS with BCT (see Table 4). 8,10,12,13 In these studies the proportion of patients found to have a positive PPM ranged from 2%24%. PPM led to a change in surgical management in 2%19% of patients. However, only 2%-9% of patients were found to have residual malignancy on re-excision performed exclusively because of the PPM findings (ie, there was no other rationale for re-excision such as close or positive margins). Although the findings of the current study are similar to those of these prior publications, the current study had a much larger sample size and was restricted to patients with DCIS. Our study has several limitations. First, our results are subject to selection bias. As PPMs were not routinely performed in all patients, it is likely that PPMs were preferentially obtained in patients with a higher likelihood of having residual disease in the breast after initial excision (eg, patients with close or positive margins). Although we did not observe an improvement in LRFS in patients undergoing PPM, it is possible the imbalance of adverse prognostic factors in the PPM group obscured what would have been a statistically significant improvement in LRFS in this group. Second, our database lacked complete information on some variables, which may have influenced the likelihood of obtaining a positive PPM such as the extent of calcifications on the initial mammogram, the use of specimen radiography, and the findings of specimen radiography. Third, the test characteristics of PPM
e112
J.T. Whaley et al
(sensitivity, specificity, positive predictive value, and negative predictive value) cannot be calculated from our data set as re-excisions were not performed in all patients with negative PPMs.
Practical Radiation Oncology: July-September 2013
2.
3.
Conclusions PPM appears to result in a change in surgical management in a small percentage of patients treated for DCIS with BCT and excision of additional malignancy in an even smaller proportion. Furthermore, PPM does not appear to have any measurable impact on LRFS. The reasons underlying the apparently modest impact of PPM are clear. When the surgical margins are close or positive or the specimen radiograph does not clearly demonstrate the suspicious calcifications seen on the screening or diagnostic mammograms, re-excision is likely to be performed regardless of the results of a PPM. Conversely, when the surgical margins are negative and the specimen radiograph identifies the calcifications of interest, PPM is unlikely to be positive. Furthermore, in the small subset of these patients who would be found to have RSC on a PPM, whole breast radiation therapy may well be sufficient to sterilize any minute amounts of residual disease associated with these microcalcifications. It is well established, for example, that the proportion of patients found to have multifocal disease on breast magnetic resonance imaging (16%) far exceeds the observed local recurrence rate following BCT. 14 In institutions where careful assessment of pathologic margins is performed and specimen radiography is routinely performed, the current study suggests that the routine PPM is not warranted.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
References 1. Bijker N, Meijnen P, Peterse JL, et al. Breast-conserving treatment with or without radiotherapy in ductal carcinoma-in-situ: ten-year results of European Organisation for Research and Treatment of Cancer randomized phase III trial 10853–a study by the EORTC
14.
Breast Cancer Cooperative Group and EORTC Radiotherapy Group. J Clin Oncol. 2006;24:3381-3387. Fisher B, Dignam J, Wolmark N, et al. Lumpectomy and radiation therapy for the treatment of intraductal breast cancer: findings from National Surgical Adjuvant Breast and Bowel Project B-17. J Clin Oncol. 1998;16:441-452. Holmberg L, Garmo H, Granstrand B, et al. Absolute risk reductions for local recurrence after postoperative radiotherapy after sector resection for ductal carcinoma in situ of the breast. J Clin Oncol. 2008;26:1247-1252. Macdonald HR, Silverstein MJ, Lee LA, et al. Margin width as the sole determinant of local recurrence after breast conservation in patients with ductal carcinoma in situ of the breast. Am J Surg. 2006;192:420-422. Solin LJ, Fourquet A, Vicini FA, et al. Mammographically detected ductal carcinoma in situ of the breast treated with breast-conserving surgery and definitive breast irradiation: long-term outcome and prognostic significance of patient age and margin status. Int J Radiat Oncol Biol Phys. 2001;50:991-1002. Solin LJ, Fourquet A, Vicini FA, et al. Long-term outcome after breastconservation treatment with radiation for mammographically detected ductal carcinoma in situ of the breast. Cancer. 2005;103:1137-1146. Morrow M, Strom EA, Bassett LW, et al. Standard for the management of ductal carcinoma in situ of the breast (DCIS). CA Cancer J Clin. 2002;52:256-276. Aref A, Youssef E, Washington T, et al. The value of postlumpectomy mammogram in the management of breast cancer patients presenting with suspiciouis microcalcifications. Cancer J Sci Am. 2000;6:25-27. Gluck BS, Dershaw DD, Liberman L, Deutch BM. Microcalcifications on postoperative mammograms as an indicator of adequacy of tumor excision. Radiology. 1993;188:469-472. Grann A, Abdou JC, Dragman N, Goodman R. The value of postexcision preradiation mammography in patients with early-stage breast cancer. Am J Clin Oncol. 2004;27:285-288. Teixidor HS, Chu FC, Kim YS, Levin TL. The value of mammography after limited breast surgery and before definitive radiation therapy. Cancer. 1992;69:1418-1423. Waddell BE, Stomper PC, DeFazio JL, Hurd TC, Edge SB. Postexcision mammography is indicated after resection of ductal carcinoma-in-situ of the breast. Ann Surg Oncol. 2000;7: 665-668. Kestin LL, Goldstein NS, Martinez AA, et al. Mammographically detected ductal carcinoma in situ treated with conservative surgery with or without radiation therapy: patterns of failure and 10-year results. Ann Surg. 2000;231:235-245. Houssami N, Ciatto S, Macaskill P, et al. Accuracy and surgical impact of magnetic resonance imaging in breast cancer staging: systematic review and meta-analysis in detection of multifocal and multicentric cancer. J Clin Oncol. 2008;26:3248-3258.
