| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Breast Imaging |
1 From the Departments of Radiology (M.A.H., J.E.B., M.A.R.), Surgery (H.A.P., A.E.C., E.G.W.), and Radiation Oncology (L.J.P.), University of Michigan Health System, 1500 E Medical Center Dr, Taubman Center 2910N, Ann Arbor, MI 48109-0326. From the 1999 RSNA scientific assembly. Received November 29, 2000; revision requested January 4, 2001; revision received September 24; accepted February 1, 2002. Address correspondence to M.A.H.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: During a 25-month study period, 214 consecutive screening mammograms in 113 asymptomatic women (mean age, 51 years) with TRAM flap reconstructions were obtained. Mastectomies were performed for cancer in 106 (94%) of the 113 women and for prophylaxis in seven (6%). Prospectively, a Breast Imaging Reporting and Data System (BI-RADS) assessment category 1–5 was assigned to each mammogram. Surgical, medical, pathologic, and radiographic records were retrospectively reviewed. CIs were determined by the normal approximation to the binomial distribution.
RESULTS: Seven (3%) of 214 examinations were BI-RADS category 4 or 5. Six (86%) of seven patients underwent biopsy. Two (33%) of these six biopsies demonstrated invasive ductal carcinoma. Cancer detection rate for mammography was 1.9% (two of 106) (95% CI: 0.33%, 7.32%) for women with reconstruction for breast cancer during the 2-year period. One (6%) of 16 BI-RADS category 3 examinations later proved to be invasive ductal carcinoma at follow-up. No interval cancer was discovered in 171 cases of BI-RADS category 1 or 2 examinations with 1-year follow-up. No cancers occurred in women who underwent prophylactic mastectomy. A biopsy positive predictive value of 33% (95% CI: 6%, 76%) was observed.
CONCLUSION: Screening mammography of TRAM flap–reconstructed breasts enables detection of nonpalpable cancer before clinical examination.
© RSNA, 2002
Index terms: Breast neoplasms, 00.324 • Breast neoplasms, postoperative, 00.324 • Breast neoplasms, radiography, 00.111, 00.113, 00.114 • Breast radiography, 00.111, 00.113, 00.114
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Recurrent breast cancer following reconstruction, although uncommon, has been reported (2–10), with 4%–11% of patients developing recurrence in the TRAM-reconstructed breast, most within the first 5 years. Annual incidence of 1% has been reported (9). Because widespread use of TRAM reconstruction is relatively new, the true long-term incidence will likely not be known for some time. Annual local recurrence rates following mastectomy without reconstruction range from 0.2% to 1% (11). The biologic and clinical importance of TRAM recurrence is incompletely understood. TRAM recurrence may be a manifestation of systemic disease or may represent the only focus of invasive carcinoma. Therefore, the role of early detection of local recurrence is not clear, and further studies will be necessary to assess the effect on morbidity and survival.
Surveillance of the TRAM-reconstructed breast for cancer detection often has been performed with regular clinical breast examination (6,12,13), although some sites routinely perform mammography as well. Those who use clinical examination alone believe that physical examination allows early detection and note the paucity of studies in which the efficacy of mammographic screening of TRAM recurrence (12) is assessed. In contrast, surveillance following breast-conserving therapy (BCT) is performed with both annual mammography and regular clinical breast examination. The annual incidence (1%–2%) of recurrence after BCT is similar to that of TRAM recurrence (14). Because mammography enables detection of nonpalpable recurrence at a smaller size than does clinical examination, screening patients with TRAM reconstruction after mastectomy for cancer may be reasonable. Screening may be especially relevant in those patents who have undergone mastectomy and TRAM flap reconstruction for early stage carcinoma without metastatic disease.
The purpose of our study was to evaluate findings from routine mammographic screenings in patients with TRAM flap reconstructions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Patients who underwent their primary surgical treatment at our institution, a National Cancer Institute–designated Comprehensive Cancer Center, were treated with mastectomy by a surgical oncologist and with TRAM flap reconstruction by a plastic surgeon. Patients’ ages at time of mastectomy ranged from 29 to 66 years (median age, 44 years; mean age, 46 years). Reconstruction occurred either at the time of mastectomy or at a later date. Patients who underwent surgical treatment outside our institution, but were being followed up by an oncologist at this institution, were also included. Surgical, medical, pathologic, and radiographic records were retrospectively reviewed, with data entered into a standard database (Excel; Microsoft, Seattle, Wash). Data reviewed (J.E.B., M.A.H.) were patient age, initial biopsy report, subsequent biopsy report, date of TRAM, date of mammography, prospective mammographic assessment, ultrasonographic (US) findings, duration of follow-up, and disease status at follow-up (until March 2001).
Mammography, Interpretation, and Biopsy
Mammography was performed following the standard technique used for nonreconstructed breasts. Initial routine screening mammography of TRAM-reconstructed breasts was a three-view study—craniocaudal, mediolateral oblique, and lateral views—with use of the same institutional protocol as that for the first screening mammographic examination for patients who underwent interval surgical biopsy. Subsequent yearly screening mammograms of the TRAM-reconstructed breast were obtained by using a two-view technique—craniocaudal and mediolateral oblique views. Mammography equipment (DMR; GE Medical Systems, Milwaukee, Wis) was set in the "contrast" mode, unless thickness exceeded 6 cm in which case the "standard" mode was used. The site was a Mammography Quality Standards Act– and State-certified facility. Study findings were interpreted by a breast radiologist, who fulfilled Mammography Quality Standards Act requirements to interpret mammograms, while the patient remained in the department. Any additional views or breast US scans were obtained at the time of the initial visit.
All mammograms had been prospectively assessed by using the American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) (15): category 1, negative; 2, benign; 3, probably benign; 4, suspicious; and 5, highly suggestive of a malignancy. Patients with category 1 or 2 mammograms were recommended for routine annual mammography. Patients with category 3 lesions were recommended for short-term mammographic follow-up at 4–6 months. Patents with category 4 or 5 lesions were recommended for biopsy.
Biopsy of BI-RADS categories 4 and 5 lesions was performed by means of image-directed core-needle biopsy or wire localization biopsy. Fine-needle aspiration was recommended for lesions believed to represent complex cysts or for lesions in which core biopsy was technically impossible. Image-guided biopsy techniques were similar to those of native breasts. The major vascular supply to the TRAM flap was known before biopsy because, TRAM reconstruction may be performed as a pedicle flap, with the inferior epigastric blood vessels coming from the lower inner quadrant, or as a free flap, with inferior epigastric or thoracodorsal anastomosis entering from the upper outer quadrant (13,16). In each biopsy case, we consulted with the surgeon to ensure that there would be no risk of traversing the TRAM vascular supply via the biopsy approach chosen.
Mammographic technique in patients with TRAM reconstructions was not substantially different from that of routine screening of native breasts or breasts after lumpectomy. The compressed thickness of TRAM-reconstructed breasts was greater than normal, with a mean thickness of 6.3 cm. However, because most of the tissue was fatty, this did not create a problem with penetration. Mammographic technical factors recorded on the patient’s mammogram, including kilovolt peak, milliampere second, and compressed breast thickness (craniocaudal view), were tabulated for each patient’s examination. The posterior nipple line length on the craniocaudal view was measured by hand retrospectively.
Breast US (model 700; GE Medical Systems) with 7–12-MHz transducers was performed by the radiologist to characterize a mammographic mass. Results were categorized as a simple cyst, complex cyst, solid mass, or negative.
Statistical Analysis
True-positive assessments were defined as a mammographic assessment category 4 or 5 and a histologic diagnosis of carcinoma. Lobular carcinoma in situ was not considered a carcinoma for this analysis. False-positive findings were defined as category 4 or 5 mammograms with a benign histologic diagnosis. False-negative findings were defined as a mammographic assessment 1, 2, or 3 with a cancer diagnosis at histologic examination within 1 year. A true-negative finding was interpreted as a category 1, 2, or 3 mammographic assessment with negative histologic diagnosis or normal mammogram within 1 year. The 95% CIs were determined by the normal approximation to the binomial distribution.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Of six patients who underwent biopsy for BI-RADS category 4 or 5 lesions, two (33%) had invasive ductal carcinoma at histologic examination (Figs 1, 2). Benign histologic findings in four (67%) of the six patients who underwent biopsy included two epidermoid inclusion cysts, one case of fat necrosis, and one case of fibrocystic change with calcifications (Fig 3). Biopsy was not performed in one patient with suspicious microcalcifications on mammograms because she was found to have metastatic disease.
|
|
|
|
|
Follow-up information was available for all 16 BI-RADS category 3 mammograms for a median time of 24 months (range, 12–36 months). Thirteen (81%) of these 16 cases were stable at mammographic follow-up and considered benign. In the other three cases (19%), the patient underwent biopsy. One patient had invasive ductal ductal cancer (1.2 cm) discovered at 6 months, and two patients had benign biopsy results.
The mammographic cancer detection rate for the 106 patients with prior cancer as the reason for mastectomy and TRAM reconstruction was 1.9% (two of 106 patients) (95% CI: 0.33%, 7.32%) during the 2-year study period. One additional cancer was later found in a patient with a BI-RADS category 3 finding at follow-up and was considered a false-negative mammographic finding. The positive predictive value for cancer detection for mammographic lesions recommended for biopsy was 33% (95% CI: 6%, 76%). By excluding the 20 BI-RADS category 1 or 2 examinations without documented follow-up and the single category 4 examination without biopsy, there were two true-positive findings, four false-positive findings, one false-negative finding, and 186 true-negative findings. Therefore, sensitivity was 67% and specificity was 98%.
Mammographic features in the two patients who underwent biopsy and in whom carcinoma was diagnosed during the study period were one mass with irregular margins adjacent to the pectoral muscle and one mass with spiculated margins in the extreme upper outer quadrant of the breast (Figs 1, 2). Both patients had an initial histologic diagnosis of ductal carcinoma in situ (DCIS) in association with invasive ductal carcinoma. Mammographic appearance of the BI-RADS category 3 tumor was focal asymmetric density on the first mammogram. The initial histologic diagnosis was invasive ductal cancer. Two of the three patients underwent immediate reconstruction at the time of original mastectomy. All patients were treated with lumpectomy and radiation therapy after discovery of the recurrence. Size of the recurrent invasive cancers at pathologic examination was 12–13 mm (mean, 12.7 mm). The time from the original diagnosis of cancer to diagnosis of TRAM recurrence was 6–10 years (mean, 7.3 years). Owing to the long duration from original diagnosis, these cancers could represent second primaries or recurrences.
Additional mammographic views were required for 36 (17%) of the 214 examinations. US was used for four examinations (2%). All four US examinations revealed solid masses (three benign, one malignant) for which biopsy was performed. Five of six biopsies were performed by means of image-directed needle biopsy and the remaining one, by means of wire localization. In the craniocaudal view, mean technical factors were 27 kVp, 156 mAs, and 6.3-cm compressed breast thickness. The mean posterior nipple line length on the craniocaudal view was 7.7 cm.
Initial histologic findings at mastectomy for 109 breasts with cancer (106 patients) were as follows: DCIS in 25 (23%); DCIS and invasive ductal in 40 (37%); invasive ductal in 16 (15%); invasive lobular in 10 (9%); and other cancer in 18 (17%).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Although mastectomy is generally believed to be a definitive procedure, local recurrence in the chest wall or in the reconstructed breast does infrequently occur (2–10,18). Chest-wall recurrence after mastectomy is reported in 0.2%–1% of women per year (11). Recurrent local disease following mastectomy and reconstruction is similar to local recurrence following mastectomy alone. Singletary (6) reported a 4.2% recurrence rate in 95 patients with TRAM reconstructions who were followed up for more than 4 years. Kroll et al (9) noted a 6.7% TRAM recurrence rate at 5.6 years in 104 patients treated with skin-sparing mastectomy and reconstruction. Noone et al (7) reported a 6.7% TRAM recurrence rate at 6.4 years, and Johnson et al (8) reported a 6% local recurrence rate in 118 women with reconstructions (primarily implants) at 2.3 years. Slavin et al (10) reported an 11.7% recurrence rate at 5.4 years in 120 patients treated with TRAM or latissimi dorsi reconstructions, but 12% of the population included patients with stage 3 or 4 lesions. Overall, an annual recurrence rate of 1%–2% has been reported for T1 and T2 tumors during the first 5 years.
Routine mammographic imaging is technically limited in those patients who have undergone reconstruction with prosthetic implants and has not generally been advocated. The situation with TRAM reconstructions is different. Because abdominal adipose tissue forms the bulk of the TRAM reconstruction, mammography of the TRAM flap is technically feasible (5,6,19–21). Opinions differ regarding the need for mammography in these women (5,6,9,19–21). This practice standard is different than the recommendation for women undergoing BCT, in whom routine mammography of the treated breast is advocated and widely practiced (22). The local recurrence rate in women undergoing BCT with lumpectomy and radiation therapy is 1%–2% per year (14), similar to the recurrence rate following TRAM reconstruction (6,7,9,10).
On the basis of small series in which the mammographic appearance of recurrent carcinoma in TRAM flaps and imaging findings in TRAM reconstructions (5,12,16,19–21) were described, physicians in our breast cancer center became convinced that early mammographic detection of cancer recurrence in TRAM flap reconstructions was technically possible. The potential benefit of screening TRAM reconstructions would be the detection of recurrences at a smaller size than appreciated clinically. Early detection of recurrent disease would likely be most beneficial to women with early stage carcinoma without evidence of metastatic disease. Potential problems from screening TRAM reconstructions include the generation of false-positive assessments, a low biopsy positive predictive value, and the unproved benefit of early detection of recurrence.
In breasts reconstructed with the TRAM flap technique, we found that screening mammography indeed enables detection of nonpalpable recurrences before clinical detection. The mammographic cancer detection rate was 1.9% (two of 106 patients) (95% CI: 0.33%, 7.32%) during the 2-year study period. The overall cancer recurrence rate (includes single false-negative category 3 lesion) was 2.8% (three of 106 patients). These rates are similar to the palpable 1% annual recurrence rate previously reported in patients with reconstructions (6,9,10) whose recurrence was detected at clinical examination. Of note, all recurrences in our study were discovered more than 5 years after mastectomy, which shows that recurrence continues to be a problem even after this interval.
One-third of mammographic abnormalities for which biopsy was performed in our series were shown to represent carcinoma. Our series is too small and CIs too great to determine if this positive predictive value will be similar to that of routine screening mammography and consistent with that of clinical practice guidelines (23). False-positive mammographic findings occurred. Benign epidermoid inclusion cysts, which can manifest as growing masses that arise as a consequence of the TRAM operation, were the most common false-positive findings. Although fat necrosis may occur in 2%–13% of TRAM-reconstructed breasts (13), fat necrosis necessitating biopsy was unusual, occurring in one of 214 examinations. A single false-negative mammographic finding (category 3) occurred during a first mammographic examination. A larger series will be needed to more accurately determine the true false-negative rate and positive predictive value.
Extensive DCIS and invasive cancer were present in the original malignancy in two of three patients with recurrence. A high frequency of extensive DCIS as the original malignancy was also found in a previous study of TRAM recurrence (5). Extensive DCIS may be associated with higher risk owing to the extensive geographic location of the disease. Women with extensive DCIS are generally not eligible for breast conservation and form a disproportionately high percentage of patients undergoing mastectomy. Most patients (59%) in our study had DCIS with or without invasive cancer. Hence, case selection bias would predict that most recurrences would also be in this group.
Although screening the TRAM-reconstructed breast can help detect nonpalpable carcinoma, early detection of TRAM recurrence may or may not be associated with improved survival. TRAM recurrence in patients with invasive carcinoma may merely be a manifestation of systemic disease and aggressive tumor biology. Early detection of in-breast recurrence in patients treated with BCT is associated with longer survival, but lead-time bias may explain this finding (24). Early in-TRAM recurrence detection would be most important in patients with DCIS alone (no invasive cancer) and in those without metastatic disease. Even if survival is not affected, morbidity may be improved as diagnosis of early in-breast recurrence may allow for better local control. All patients in our series were treated with lumpectomy and radiation therapy. Clinical standards of care in the United States include aggressive surveillance with mammography and physical examination for patients treated with BCT. It may be reasonable to apply the same practice standards to women treated with TRAM reconstruction.
Why breast cancer recurs in TRAM flaps has important biologic ramifications for understanding tumor behavior. Postulated mechanisms include residual cancer tissue, tumor seeding at time of mastectomy, sequestration of tumor cells within lymphatics, and unspecified host factors. Late "recurrence" could also be a new primary tumor that arises from benign breast tissue left after mastectomy. Mastectomy removes most (>95%) breast cells at risk but does not remove all breast cells at risk (21). This is well demonstrated in Figure 3, the case of a patient with extreme upper outer quadrant calcifications that were identified as fibrocystic change with calcifications at biopsy. In this patient, these fibrocystic changes were likely in breast tissue not excised at the time of mastectomy.
Should routine screening of TRAM flaps be advocated for all patients? Our series is too small to make a definitive recommendation, but our findings provide inferential support of the practice. We continue to perform mammography in patients with TRAM reconstructions for cancer. Cancer occurrence in women treated with prophylactic mastectomy and TRAM reconstruction was not found. Because a small amount of residual breast tissue persists after prophylactic mastectomy, a small risk for cancer exists (25). The decision to mammographically screen this group of patients is best made following a balanced discussion between the patient and her physician.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, M.A.H., J.E.B.; study concepts and design, all authors; literature research, M.A.H., L.J.P., J.E.B., M.A.R., H.A.P.; clinical studies, M.A.H., J.E.B., M.A.R.; data acquisition, M.A.H., J.E.B., M.A.R., H.A.P., L.J.P., A.E.C.; data analysis/interpretation, all authors; manuscript preparation, definition of intellectual content, editing, revision/review, and final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. M. Lee, D. Georgian-Smith, G. S. Gazelle, E. F. Halpern, E. A. Rafferty, R. H. Moore, E. D. Yeh, H. A. D'Alessandro, R. A. Hitt, and D. B. Kopans Detecting Nonpalpable Recurrent Breast Cancer: The Role of Routine Mammographic Screening of Transverse Rectus Abdominis Myocutaneous Flap Reconstructions Radiology, August 1, 2008; 248(2): 398 - 405. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C GLYNN and J LITHERLAND Imaging breast augmentation and reconstruction Br. J. Radiol., July 1, 2008; 81(967): 587 - 595. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. P. Deguzman, K. L. Bui, J. J. Lenert, J. A. Rapelyea, and R. F. Brem Mammographic Appearance of Lymphedema in a TRAM-Reconstructed Breast Am. J. Roentgenol., December 1, 2004; 183(6): 1810 - 1812. [Full Text] [PDF]
|
|
|
|
 |
|
 R. K. Devon, M. A. Rosen, C. Mies, and S. G. Orel Breast Reconstruction with a Transverse Rectus Abdominis Myocutaneous Flap: Spectrum of Normal and Abnormal MR Imaging Findings RadioGraphics, September 1, 2004; 24(5): 1287 - 1299. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Kim and J. M. Park Mammographic and Ultrasonographic Features After Autogenous Myocutaneous Flap Reconstruction Mammoplasty J. Ultrasound Med., February 1, 2004; 23(2): 275 - 282. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
