Wo m e n ’s I m a g i n g • O r i g i n a l R e s e a r c h Chae et al. Breast MRI of Residual Disease

Downloaded from www.ajronline.org by 37.44.207.16 on 01/22/17 from IP address 37.44.207.16. Copyright ARRS. For personal use only; all rights reserved

Women’s Imaging Original Research

Evaluation of Residual Disease Using Breast MRI After Excisional Biopsy for Breast Cancer Eun Young Chae1 Joo Hee Cha1 Hak Hee Kim1 Hee Jung Shin1 Hyunji Kim1 JungBok Lee 2 Joo Yeon Cheung1 Chae EY, Cha JH, Kim HH, et al.

Keywords: breast cancer, excisional biopsy, MRI, residual disease DOI:10.2214/AJR.12.9275 Received May 17, 2012; accepted after revision August 9, 2012. 1

Department of Radiology and Research Institute of Radiology, University of Ulsan College of Medicine, Asan Medical Center, 388-1 Pungnap-2-dong, Songpa-gu, Seoul 138-736, Korea. Address correspondence to J. H. Cha ([email protected]). 2 Department of Clinical Epidemiology and Biostatistics, Asan Medical Center, College of Medicine, University of Ulsan, Seoul, Korea.

AJR 2013; 200:1167–1173 0361–803X/13/2005–1167 © American Roentgen Ray Society

OBJECTIVE. The purpose of our study was to assess the diagnostic performance of breast MRI in evaluating residual disease in patients after excisional biopsy for breast cancer on the basis of both morphologic and kinetic features. MATERIALS AND METHODS. Of 5304 breast MRI examinations performed between January 2007 and December 2011, 308 evaluated postoperative sites after excisional biopsy of breast cancer. Among these, 203 patients who were not treated with chemotherapy or radiotherapy before MRI and underwent definitive surgery within 30 days after MRI were enrolled. MRI findings were analyzed on the basis of contrast-enhanced subtraction images. The enhancement patterns were classified into four categories: no enhancement (P1), thin regular rim enhancement (P2), thick or irregular rim enhancement (P3), and nodular or nonmasslike enhancement (P4) around the postoperative sites. The enhancement kinetics were assessed as follows: persistent, plateau, and washout pattern. RESULTS. From 207 breast MRI examinations in 203 patients, 144 breasts had residual breast cancer at histopathologic examination after definitive surgery. When P1 and P2 were considered negative for residual cancer and P3 and P4 were considered positive, the sensitivity, specificity, positive and negative predictive values, and accuracy were 79.9%, 73.0%, 87.1%, 61.3%, and 77.8%, respectively. The specificity and positive predictive value improved to 90.5% and 91.7%, when analyzed with washout enhancement kinetics as another positive finding for residual cancer. A statistically significant trend of decreasing specificity and positive predictive value (p < 0.05) was found with the passage of a time interval between excision and breast MRI. CONCLUSION. Although the overlapping features of the postsurgical changes and malignant lesions remain as the limitations, dynamic contrast-enhanced breast MRI is a useful tool for residual disease prediction after excisional biopsy for breast cancer. Combined use of morphologic and kinetic evaluation parameters improved the diagnostic performance. We do not recommend that MRI be unreasonably delayed after excisional biopsy considering the risk of prolonging definitive surgery.

T

he incidence of residual disease after initial excisional biopsy of breast cancer is reported to range from 45% up to 70% [1, 2]. When a biopsy is performed at another hospital, the surgical margin status is often not available. A positive margin is associated with an increased risk of local recurrence [3, 4] and is the basis for reexcision in a patient with questionable or inadequate margin status. Dynamic contrast-enhanced MRI of the breast has become an important tool in patients with known or suspected breast cancer, with consistently high sensitivity of 89–100% and more variable specificity of 26–91% [5–8]. Breast MRI is also helpful to

determine the presence or absence of residual cancer, which may aid in surgical planning for reexcision, and may identify those patients who would ultimately require mastectomy by revealing additional foci of cancer. Several previous studies have reported the role of MRI in evaluating patients who have undergone excisional biopsy for breast cancer [9–12]. However, these studies mainly focused on the morphologic analysis of postoperative sites. Little is known about the utility of kinetic evaluation of breast MRI in predicting residual disease. The purpose of this study was to assess the diagnostic performance of breast MRI in the evaluation of residual disease in pa-

AJR:200, May 2013 1167

Downloaded from www.ajronline.org by 37.44.207.16 on 01/22/17 from IP address 37.44.207.16. Copyright ARRS. For personal use only; all rights reserved

Chae et al. tients after excisional biopsy for breast cancer on the basis of both morphologic and kinetic features. Additionally, we investigated whether the time interval between initial excisional biopsy and breast MRI may influence the diagnostic performance for detecting residual disease. Materials and Methods Patient Selection This retrospective study was approved by the institutional review board, and the requirement to obtain informed consent was waived. Between January 2007 and December 2011, MRI of 5304 breasts was performed at our institution. Among these, MRI of 308 breasts was performed for the evaluation of postoperative sites after excisional biopsy of breast cancer. The patients initially underwent surgery not knowing whether they had cancer or not. After initial surgery, they were diagnosed with breast cancer, and had been scheduled for secondstep surgery. MRI was performed before definitive surgery. Inclusion in this study was based on the following criteria: the patient had breast cancer diagnosed by excisional biopsy; the patient had undergone definitive surgical treatment within 30 days after MRI; and no chemotherapy or radiotherapy was administered before MRI and definitive surgery. Because most of our patients underwent primary surgery at an outside hospital, the information on margin status was not available. Therefore, the margin status was not involved in the inclusion criteria. After excluding 105 patients who underwent radiotherapy or neoadjuvant chemotherapy before definitive surgery, 207 breast MRI examinations in 203 patients (age range, 22–77 years; mean age, 47 years) constituted our study population, including four patients with bilateral breast cancer.

MRI Technique All patients were scanned on a 1.5-T scanner with a bilateral breast array coil (Magnetom Avanto, Siemens Healthcare). The standard MRI protocol included the following pulse sequences: axial 2D T2-weighted STIR turbo spin-echo pulse sequence (TR/TE, 6700/74; inversion time, 150 ms; FOV, 300 × 300 mm; matrix, 448 × 448; and slice thickness, 5 mm); unenhanced and contrast-enhanced fat-saturated axial 3D T1-weighted FLASH volume-interpolated breath-hold examination pulse sequences (TR/TE, 5.2/2.4; FOV, 340 × 340 mm; matrix, 384 × 384; and slice thickness, 0.9 mm); and axial 3D delayed contrast-enhanced turbo spinecho pulse sequence (TR/TE, 767/12; FOV, 350 × 350 mm; matrix, 768 × 768; and slice thickness, 5 mm) for the evaluation of the supraclavicular and axillary lymph nodes. The six dynamic sequences were performed before and after IV injection of the

1168

contrast medium. The contrast medium (0.2 mL/kg body weight) (gadopentetate dimeglumine, Magnevist, Schering) was injected using an MR-compatible power injector (Spectris, Medrad) with a flow of 1 mL/s followed by a 20-mL saline flush. Postprocessing manipulation included the production of standard subtraction, reverse subtraction, and maximum-intensity-projection images.

MRI Interpretation MRI was retrospectively reviewed by two radiologists experienced in the interpretation of breast MRI. Each radiologist was blinded to the readings of the other radiologist during the initial review. When there was a discrepancy, the two radiologists reviewed these cases together and reached a consensus. The radiologists were blinded to any clinical or histopathologic information of the patients. The presence of residual disease was assessed by a combination of morphologic and kinetic features of the postoperative sites. The morphologic pattern was evaluated by visual observation of residual enhancement on the contrast-enhanced subtraction images and was classified into four categories: no enhancement (P1 and K1), thin regular rim enhancement of the seroma cavity (P2), thick or irregular rim enhancement of the seroma cavity (P3), and nodular or nonmasslike enhancement around the postoperative sites (P4). The enhancement kinetics were assessed based on the following patterns: persistent (K2), plateau (K3), and washout (K4). The time-signal intensity curves obtained using commercial software (CADstream, Merge Healthcare) in the area of greatest or most homogeneous enhancement or in areas where visible residual enhancement was detected were used to help differentiate residual tumor from adjacent breast parenchyma. By definition, a “persistent” pattern shows continuously increasing enhancement throughout the dynamic series. A

“plateau” pattern is an initial increase in signal intensity, which is followed by a relatively constant value throughout the delayed phase. A “washout” pattern reaches a peak at the end of the initial phase, and then the enhancement declines throughout the delayed phase [13]. Each lesion was assigned to an enhancement pattern that was most indicative of the malignancy over the entire lesion; the patterns in the order of decreasing strength were washout, followed by plateau and then persistent.

Histopathology and Its Correlation With Breast MRI The clinical and pathologic data were retrospectively reviewed. Data recorded included initial pathology results of excisional biopsy and time interval between initial biopsy and MRI evaluation. At the time of definitive surgery, the following data of the specimens were recorded: the site of initial excision, presence or absence of tumor at the margin, and location and size of residual disease on the basis of histopathologic evaluation. The MRI data were then correlated with the final histopathologic findings.

Statistical Analysis All the statistical calculations were performed using SPSS software (version 13.0, Statistical Package for the Social Sciences). We calculated the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of each MRI finding for the prediction of residual breast cancer. We also performed a trend test to evaluate the effect of the time interval between excision and MRI on the diagnostic performance of breast MRI to predict residual breast cancer. The time interval was classified into four categories: within 7 days, between 8 and 14 days, between 15 and 21 days, and 22 or more days. The trend test was a regression analysis, for which the

TABLE 1:  Initial Histopathologic Diagnosis Diagnosis

No.

%

Invasive ductal carcinoma, not otherwise specified

89

43.0

Ductal carcinoma in situ

99

47.8

Invasive lobular carcinoma

5

2.4

Mucinous carcinoma

5

2.4

Other

Tubular carcinoma

3

1.4

Intraductal papillary carcinoma

3

1.4

Mixed lobular and ductal carcinoma

1

0.5

Intracystic papillary carcinoma

1

0.5

1

0.5

Lobular carcinoma in situ Total

207

AJR:200, May 2013

Breast MRI of Residual Disease TABLE 2:  Diagnostic Performance of Each MRI Characteristic Characteristic

Downloaded from www.ajronline.org by 37.44.207.16 on 01/22/17 from IP address 37.44.207.16. Copyright ARRS. For personal use only; all rights reserved

Enhancement pattern

Enhancement kinetics

No.

Sensitivity (%)

Specificity (%)

PPV (%)

NPV (%)

Accuracy (%)

P1 (= K1)

39

10.42

61.90

38.46

23.21

26.09 26.57

P2

36

9.72

65.08

38.89

23.98

P3

60

34.03

82.54

81.67

35.37

48.79

P4

72

45.83

90.48

91.67

42.22

59.42

K2

77

34.03

55.56

63.64

26.92

40.58

K3

19

9.72

92.06

73.68

30.85

34.78

K4

72

45.83

90.48

91.67

42.22

59.42

Note—P1 and K1 = no enhancement, P2 = thin regular enhancement of the seroma cavity, P3 = thick or irregular rim enhancement of the seroma cavity, P4 = nodular or nonmasslike enhancement around the postoperative sites, K2 = persistent, K3 = plateau, K4 = washout, PPV = positive predictive value, NPV = negative predictive value. Fig. 1—38-year-old woman with invasive ductal carcinoma. A and B, Axial dynamic contrast-enhanced subtracted T1-weighted FLASH volume-interpolated breath-hold examination images obtained 2 minutes (A) and 5 minutes (B) after injection of contrast medium show thin regular rim enhancement (arrows, P2) around seroma cavity in right breast. Timesignal intensity curve exhibits persistent pattern (K2). Pathology after definitive surgery revealed no residual cancer.

statistical inference was derived using the slope estimate and the corresponding Student t statistic. A p value of less than 0.05 was considered to indicate a statistically significant difference.

Results The histopathologic findings for the diagnosis of initial excisional biopsy are summarized in Table 1. Of 207 breasts in 203 patients, 144 breasts (70.9%) had residual breast cancer at histopathologic examination after definitive surgery. The MRI findings are summarized in Table 2. The patterns of enhancement of the postexcisional sites were as follows: no enhance-

Fig. 2—58-year-old woman with invasive ductal carcinoma. A and B, Axial dynamic contrast-enhanced subtracted T1-weighted FLASH volume-interpolated breath-hold examination images obtained 2 minutes (A) and 5 minutes (B) after injection of contrast medium show thick and irregular rim enhancement (arrows, P3) around seroma cavity in right breast. Time-signal intensity curve exhibits washout pattern (K4). Pathology after definitive surgery revealed residual invasive ductal carcinoma.

A

B

ment (P1) in 39 breasts (18.8%), thin regular rim enhancement (P2) (Fig. 1) in 36 breasts (17.4%), thick or irregular rim enhancement (P3) (Fig. 2) in 60 breasts (29.0%), and nodular (Fig. 3) or nonmasslike (Fig. 4) enhancement around the postoperative sites (P4) in 72 breasts (34.8%). The kinetic features of enhancement were as follows: no enhancement (K1) in 39 breasts (18.8%), persistent pattern (K2) in 77 breasts (37.2%), plateau pattern (K3) in 19 breasts (9.2), and washout pattern (K4) in 72 breasts (34.8%). Table 2 shows the sensitivity, specificity, PPV, NPV, and overall accuracy for each MRI finding and for morphologic and kinetic

patterns of enhancement used to predict residual malignancy. When P1 and P2 were considered negative for residual cancer and P3 and P4 were considered positive (not shown in table), the sensitivity, specificity, PPV, NPV, and accuracy were 79.9%, 73.0%, 87.1%, 61.3%, and 77.8%, respectively. The specificity and PPV improved to 90.5% and 91.7%, respectively, when analyzed with the washout enhancement kinetics used as another positive finding for residual cancer in addition to P3 and P4 (Table 3). With these criteria, six patients had false-positive MRI findings. The pathologic results with false-positive MRI findings were intraductal papilloma in two pa-

A

B

AJR:200, May 2013 1169

Downloaded from www.ajronline.org by 37.44.207.16 on 01/22/17 from IP address 37.44.207.16. Copyright ARRS. For personal use only; all rights reserved

Chae et al.

Curve Peak: 108% (Rapid, Washout) Persistent 9% (9/0) 9% Medium

A

Plateau 61% (59/5)

Washout 27% (23/5)

0% Rapid

B

C

Fig. 3—41-year-old woman with invasive ductal carcinoma. A–C, Axial dynamic contrast-enhanced subtracted T1-weighted FLASH volume-interpolated breath-hold examination images obtained 2 minutes (A) and 5 minutes (B) after injection of contrast medium show nodular enhancement (arrow) at medial aspect of postoperative sites (P4). Time-signal intensity curve (C) exhibits washout pattern (K4). Pathology after definitive surgery revealed residual invasive ductal carcinoma.

tients, atypical ductal hyperplasia in one patient, foreign body reaction and usual ductal hyperplasia in one patient, and no diagnostic abnormality in two patients. The time interval between excision and MRI of the breast showed a statistically significant trend toward progressive decrease in specificity and PPV (p < 0.05) over time (Table 4 and Fig. 5). No statistically sig-

nificant difference was seen on sensitivity, NPV, and accuracy. Discussion The rate of residual disease after excisional biopsy of breast cancer is reported to be up to 70% [1, 2]. In our study, the prevalence of residual tumor was 70.9%, which falls within this range. Obviously, the goal in breast

TABLE 3:  Diagnostic Performance in Combination of Pattern and Kinetics of Enhancement Pattern Score

Kinetics Score

Sensitivity (%)

Specificity (%)

PPV (%)

NPV (%)

Accuracy (%)

P1 + P2

K2

7.64

73.02

39.29

25.70

27.53

P1 + P2

K3

2.08

92.06

37.50

29.15

29.47

P3 + P4

K2

26.39

82.54

77.55

32.91

43.48

P3 + P4

K3

7.64

100

100

32.14

35.74

P3 + P4

K4

45.83

90.48

91.67

42.22

59.42

Note—P1 and K1 = no enhancement, P2 = thin regular enhancement of the seroma cavity, P3 = thick or irregular rim enhancement of the seroma cavity, P4 = nodular or nonmasslike enhancement around the postoperative sites, K2 = persistent, K3 = plateau, K4 = washout, PPV = positive predictive value, NPV = negative predictive value.

A 1170

B

cancer treatment is the surgical removal of all foci of carcinoma. To determine adequate subsequent treatment, such as reexcision or mastectomy, the accurate evaluation of residual disease and the assessment of its extent are necessary for surgical planning after excisional biopsy of breast carcinoma. There is emerging evidence that breast MRI can provide important information regarding the presence of residual carcinoma after excisional biopsy. The preoperative extent of disease is better assessed using MRI that by conventional imaging techniques, such as mammography and ultrasound [14]. This is particularly true for the postoperative breast because postsurgical changes, such as architectural distortion and hematoma, may obscure or mimic malignancy [11]. Orel et al. [11] evaluated postoperative patients using MRI to assess for residual disease and reported PPV of 82% and NPV of 61%. In other studies, the reported sensitivity, specificity, PPV, and NPV of MRI for detecting

Fig. 4—41-year-old woman with ductal carcinoma in situ. Axial dynamic contrast-enhanced subtracted T1weighted FLASH volume-interpolated breath-hold examination images obtained 2 minutes (A) and 5 minutes (B) after injection of contrast medium show nonmasslike enhancement (arrows) with segmental distribution around postoperative sites (P4). Timesignal intensity curve exhibits washout pattern (K4). Pathology after definitive surgery revealed residual ductal carcinoma in situ.

AJR:200, May 2013

residual disease have been 61.2–92.1%, 29– 81%, 69–88.6%, and 54–63%, respectively [9–12, 15–17]. In the current study, we assessed the diagnostic performance of dynamic contrastenhanced MRI for evaluation of residual disease after excisional biopsy of breast carcinoma using various criteria based on the morphologic and kinetic characteristics of MRI. When P1 (no enhancement) and P2 (thin regular rim enhancement) were considered negative for residual disease, and P3 (thick or irregular rim enhancement) and P4 (nodular or nonmasslike enhancement) were considered positive, the sensitivity, specificity, PPV, NPV, and accuracy were 79.9%, 73.0%, 87.1%, 61.3%, and 77.8%, respectively. Our results, which are consistent with those of previous studies, reconfirm that the MRI findings of residual malignancy and those of postoperative changes have overlapping appearances (Fig. 6), and this limits the diagnostic performance of MRI in the evaluation for predicting residual disease. Because postoperative granulation tissue also shows enhancement, it is not easy to differentiate residual cancer from postsurgical changes. In addition to morphologic assessment, we also evaluated the kinetic characteristics of the breast. It is well known that the combination of both morphologic and kinetic breast evaluation is important [18–23]. Invasive cancers tend to show a rapid and intense uptake of contrast agent and exhibit a washout curve at the delayed phase [19]. In our study, the analysis of kinetic features of MRI improved specificity, and the PPVs improved to 90.5% and 91.7%, when analyzed with the washout enhancement kinetics as another positive finding for residual cancer. Although washout enhancement pattern is not a single reliable finding of residual cancer because of considerable overlap between benign and malignant lesions [19, 20], this pattern may

Fig. 6—39-year-old woman with ductal carcinoma in situ and false-positive findings on MRI. A and B, Axial dynamic contrast-enhanced subtracted T1-weighted FLASH volume-interpolated breath-hold examination (VIBE) images obtained 2 minutes (A) and 5 minutes (B) after injection of contrast medium show nodular enhancement (arrows) around seroma cavity (nonmasslike enhancement with segmental distribution around postoperative sites) (P4). Time-signal intensity curve (not shown) revealed washout pattern (K4). Pathology after definitive surgery revealed foreign body reaction and usual ductal hyperplasia without residual disease.

TABLE 4:  Influence of Time Interval on Diagnostic Performance of Breast MRI Time Interval (d)

Sensitivity (%)

Specificity (%)

PPV (%)

NPV (%)

Accuracy (%)

81.82

100

100

66.67

86.67

≤ 7 (n = 15) 8–14 (n = 52)

81.58

85.71

93.94

63.16

82.7

15–21 (n = 54)

75.68

70.59

84.85

57.14

74.07

≥ 22 (n = 86)

81.03

64.29

82.46

62.07

75.58

Trend test (p)

0.4662

0.0289

0.0272

0.4257

0.1120

Note—P1 and K1 = no enhancement, P2 = thin regular enhancement of the seroma cavity, P3 = thick or irregular rim enhancement of the seroma cavity, P4 = nodular or nonmasslike enhancement around the postoperative sites, K2 = persistent, K3 = plateau, K4 = washout, PPV = positive predictive value, NPV = negative predictive value. Fig. 5—Graph shows influence of time interval on diagnostic performance of breast MRI performed after excisional biopsy for breast cancer. PPV = positive predictive value; NPV = negative predictive value.

100 Diagnostic Performance (%)

Downloaded from www.ajronline.org by 37.44.207.16 on 01/22/17 from IP address 37.44.207.16. Copyright ARRS. For personal use only; all rights reserved

Breast MRI of Residual Disease

100 93.94

90

86.67

80

81.82

85.71 82.7 81.58

70 60

66.67

84.85 75.68 74.07 70.59

64.29 62.07

63.16 57.14

50 40

Sensitivity Specificity PPV NPV Accuracy

30 20 10 0

82.46 81.03 75.58

≤7d

8–14 d

15–21 d

≥ 22 d

serve as another criterion of residual disease. When evaluating breast MRI after excisional biopsy in daily practice, careful MRI assessment on the basis of both morphologic and kinetic features may have a role in improved differentiation of postsurgical changes and residual cancer. In this study, we found that MRI lesion size measurement correlated with pathologic tumor size within (no more than) twofold in 51 of 66 true-positive cases (77%) when P3 and P4 as well as washout kinetics were considered as positive findings for residual cancer. In eight of the 15 discordant cases, the size differences were primarily due to the presence

of an intraductal component around the invasive carcinoma, and the extent of the intraductal component correlated with MRI lesion size. The use of MRI in diagnosing intraductal extension of breast cancer has been reported [24], and the exact measurement of the extent of the intraductal component is important in preventing local recurrence for planning breast-conserving surgery [25]. Although some patients have discordance of MRI and pathologic cancer size, a fair number of these patients will have significant pathologic abnormalities around the main tumor. In our study, the most common pathologic diagnosis was ductal carcinoma in situ (47.8%)

A

B

AJR:200, May 2013 1171

Downloaded from www.ajronline.org by 37.44.207.16 on 01/22/17 from IP address 37.44.207.16. Copyright ARRS. For personal use only; all rights reserved

Chae et al. and not invasive ductal carcinoma (Table 1). The reported rates of ductal carcinoma in situ in previous studies are between 17% and 29.4% [9–11], which are lower than our findings. The sensitivity of MRI for the identification of ductal carcinoma in situ has been reported to be variable (40–100%) and is usually lower than that of invasive ductal carcinoma [26, 27]. Despite the higher prevalence of ductal carcinoma in situ, the sensitivity of MRI for predicting residual disease in our study was comparable with previous reports. To the best of our knowledge, there is a lack of consensus on the optimal time interval between excisional biopsy and MRI. A prior study published in 2000 [28] reported that the highest specificity of 75% for the evaluation of residual disease was reached between 28 and 35 days after surgery. The authors of the study recommended that at least 28 days should elapse before determining whether there is residual disease. In contrast, a later study [16] reported that the specificity and sensitivity for the evaluation of residual disease were higher when MRI was performed within 28 days of the original surgery. In our study, we found that there is a trend toward decreasing specificity and PPV with the passage of time between excision and breast MRI. Although the natural history of postoperative change and healing processes in brain may considerably differ from that of the breast, Forsyth et al. [29] reported that MRI performed on postoperative days 3–5 minimized the confounding effects of postsurgical enhancement for accurate assessment of residual tumor in patients with malignant gliomas. Postoperative sites may appear enhanced up to 6 months after surgery without radiation therapy and up to 18–24 months after radiation therapy [30]. Our hypothesis is that early postoperative MRI may be useful before nonneoplastic contrast enhancement from postsurgical changes becomes radiologically apparent. Therefore, we do not recommend that MRI be unreasonably delayed after excisional biopsy considering the risk of prolonging definitive surgery. In our study, 29.1% of cases had no residual disease after definitive surgery, and the surgery was not necessary for these patients. Currently, no reliable criteria are established to determine which patients must undergo reexcision and which patients can be spared further surgery. One of the issues is that some residual disease can be treated by radiation therapy or adjunctive chemotherapy that is generally recommended for patients who have had breast-conserving surgery for inva-

1172

sive breast cancer. Although a reexcision is still recommended at this time because of the overlapping features of residual cancer and postsurgical changes, it may be useful in the future to use MRI to identify those patients in whom further surgery may not be necessary. Our study has certain limitations. The patients in our study initially underwent excisional biopsy instead of core needle biopsy. The standard approach to evaluating breast abnormalities at our institution is to perform an imaging-guided core biopsy rather than a diagnostic excisional biopsy. With the use of core biopsy, the diagnosis of breast cancer could be made before the surgical procedure, and this technique can avoid the postsurgical changes that are a cause of pitfalls in postoperative MRI. Because a large number of our patients were referred from outside facilities, we cannot determine the clinical course of these patients due to the retrospective nature of this study. Furthermore, although radiologic-pathologic correlation of all MRI abnormalities would have provided more definitive information, we could not pathologically confirm the MRI abnormalities one by one. In conclusion, although postsurgical changes and the overlapping features of benign and malignant lesions remain to be the limitations, dynamic contrast-enhanced breast MRI is a useful tool for the prediction of residual disease after excisional biopsy in breast cancer patients. Moreover, the combined use of the morphologic and kinetic evaluation parameters improved the diagnostic performance of breast MRI. References 1. Jardines L, Fowble B, Schultz D, et al. Factors associated with a positive reexcision after excisional biopsy for invasive breast cancer. Surgery 1995; 118:803–809 2. Gwin JL, Eisenberg BL, Hoffman JP, Ottery FD, Boraas M, Solin LJ. Incidence of gross and microscopic carcinoma in specimens from patients with breast cancer after re-excision lumpectomy. Ann Surg 1993; 218:729–734 3. Pittinger TP, Maronian NC, Poulter CA, Peacock JL. Importance of margin status in outcome of breast-conserving surgery for carcinoma. Surgery 1994; 116:605–608; discussion, 608–609 4. Tartter PI, Kaplan J, Bleiweiss I, et al. Lumpectomy margins, reexcision, and local recurrence of breast cancer. Am J Surg 2000; 179:81–85 5. Gilles R, Guinebretiere JM, Lucidarme O, et al. Nonpalpable breast tumors: diagnosis with contrast-enhanced subtraction dynamic MR imaging. Radiology 1994; 191:625–631

6. Harms SE. Technical report of the international working group on breast MRI. J Magn Reson Imaging 1999; 10:979 7. Berg WA, Gutierrez L, NessAiver MS, et al. Diagnostic accuracy of mammography, clinical examination, US, and MR imaging in preoperative assessment of breast cancer. Radiology 2004; 233:830–849 8. Teifke A, Hlawatsch A, Beier T, et al. Undetected malignancies of the breast: dynamic contrast-enhanced MR imaging at 1.0 T. Radiology 2002; 224:881–888 9. Kim JA, Son EJ, Kim EK, Kim MJ, Kwak JY, Jeong J. Postexcisional breast magnetic resonance imaging in patients with breast cancer: predictable findings of residual cancer. J Comput Assist Tomogr 2009; 33:940–945 10. Lee JM, Orel SG, Czerniecki BJ, Solin LJ, Schnall MD. MRI before reexcision surgery in patients with breast cancer. AJR 2004; 182:473– 480 11. Orel SG, Reynolds C, Schnall MD, Solin LJ, Fraker DL, Sullivan DC. Breast carcinoma: MR imaging before re-excisional biopsy. Radiology 1997; 205:429–436 12. Soderstrom CE, Harms SE, Farrell RS Jr, Pruneda JM, Flamig DP. Detection with MR imaging of residual tumor in the breast soon after surgery. AJR 1997; 168:485–488 13. Molleran V, Mahoney MC. The BI-RADS breast magnetic resonance imaging lexicon. Magn Reson Imaging Clin N Am 2010; 18:171–185 14. Orel SG, Schnall MD, Powell CM, et al. Staging of suspected breast cancer: effect of MR imaging and MR-guided biopsy. Radiology 1995; 196:115– 122 15. Kawashima H, Tawara M, Suzuki M, Matsui O, Kadoya M. Effectiveness of dynamic MRI for diagnosing pericicatricial minimal residual breast cancer following excisional biopsy. Eur J Radiol 2001; 40:2–9 16. Stucky CC, McLaughlin SA, Dueck AC, et al. Does magnetic resonance imaging accurately predict residual disease in breast cancer? Am J Surg 2009; 198:547–552 17. Wilkinson J, Appleton CM, Margenthaler JA. Utility of breast MRI for evaluation of residual disease following excisional biopsy. J Surg Res 2011; 170:233–239 18. Morris EA. Breast MR imaging lexicon updated. Magn Reson Imaging Clin N Am 2006; 14:293– 303 19. Kuhl CK, Mielcareck P, Klaschik S, et al. Dynamic breast MR imaging: are signal intensity time course data useful for differential diagnosis of enhancing lesions? Radiology 1999; 211:101–110 20. Buadu LD, Murakami J, Murayama S, et al. Breast lesions: correlation of contrast medium en-

AJR:200, May 2013

Downloaded from www.ajronline.org by 37.44.207.16 on 01/22/17 from IP address 37.44.207.16. Copyright ARRS. For personal use only; all rights reserved

Breast MRI of Residual Disease hancement patterns on MR images with histopathologic findings and tumor angiogenesis. Radiology 1996; 200:639–649 21. Szabó BK, Aspelin P, Wiberg MK, Boné B. Dynamic MR imaging of the breast: analysis of kinetic and morphologic diagnostic criteria. Acta Radiol 2003; 44:379–386 22. Schnall MD, Blume J, Bluemke DA, et al. Diagnostic architectural and dynamic features at breast MR imaging: multicenter study. Radiology 2006; 238:42–53 23. Wang LC, DeMartini WB, Partridge SC, Peacock S, Lehman CD. MRI-detected suspicious breast lesions: predictive values of kinetic features measured by computer-aided evaluation. AJR 2009; 193:826–831

24. Ando Y, Fukatsu H, Ishigaki T, Endo T, Miyazaki M. Intramammary extension of breast cancer: diagnosis with MR mammography. Breast Cancer 1998; 5:291–300 25. Sundararajan S, Tohno E, Kamma H, Ueno E, Minami M. Role of ultrasonography and MRI in the detection of wide intraductal component of invasive breast cancer: a prospective study. Clin Radiol 2007; 62:252–261 26. Hwang ES, Kinkel K, Esserman LJ, Lu Y, Weidner N, Hylton NM. Magnetic resonance imaging in patients diagnosed with ductal carcinoma-in-situ: value in the diagnosis of residual disease, occult invasion, and multicentricity. Ann Surg Oncol 2003; 10:381–388 27. Orel SG, Mendonca MH, Reynolds C, Schnall

MD, Solin LJ, Sullivan DC. MR imaging of ductal carcinoma in situ. Radiology 1997; 202:413–420 28. Frei KA, Kinkel K, Bonel HM, Lu Y, Esserman LJ, Hylton NM. MR imaging of the breast in patients with positive margins after lumpectomy: influence of the time interval between lumpectomy and MR imaging. AJR 2000; 175:1577–1584 29. Forsyth PA, Petrov E, Mahallati H, et al. Prospective study of postoperative magnetic resonance imaging in patients with malignant gliomas. J Clin Oncol 1997; 15:2076–2081 30. Heywang SH, Hilbertz T, Beck R, Bauer WM, Eiermann W, Permanetter W. Gd-DTPA enhanced MR imaging of the breast in patients with postoperative scarring and silicon implants. J Comput Assist Tomogr 1990; 14:348–356

AJR:200, May 2013 1173