| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology, The New York Presbyterian Hospital, Strang-Cornell
Breast Center, 525 E. 68th St., New York, NY 10021.
2
Present address: Women's Diagnostic and Wellness Center, Nyack Hospital, 160
N. Midland Ave., Nyack, NY 10960.
3
Present address: Hudson Valley Radiology Associates, PLLC, 18 Squadron Blvd.,
New City, NY 10956.
4
Department of Surgery, The New York Presbyterian Hospital, Strang-Cornell
Breast Center, New York, NY 10021.
Received May 20, 1999;
accepted after revision December 22, 1999.
Address correspondence to C. S. Giess.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Retrospective review identified 53 new calcifications at the lumpectomy bed in patients with available mammograms and medical records. Breast Imaging Reporting and Data System (BI-RADS) categories were retrospectively assigned on the basis of initial prospective recommendation for yearly follow-up (category 2), 6-month follow-up (category 3), or biopsy (category 4 or 5). Outcomes were defined as benign for no recurrence at the lumpectomy bed on biopsy or follow-up and malignant if biopsy-proven at the lumpectomy bed.
RESULTS. The median rate of development after lumpectomy was 23 months (range, 2-174 months) for benign and 39 months (range, 15-112 months) for malignant calcifications. Fifteen (28%) of 53 calcifications were classified as BI-RADS category 3. Twelve (80%) of 15 were downgraded to BI-RADS category 2 at a median follow-up of 6.5 months (range, 6-16 months); none represented recurrent disease. Three (20%) of 15 were upgraded to BI-RADS category 4 at the 6-month follow-up, one despite stability (benign) and two for increasing pleomorphism (malignant). Nine (17%) of 53 calcifications were classified as BI-RADS category 4 or 5; six (67%) of the nine were malignant and three (33%) were benign at biopsy. Twenty-nine (55%) of 53 calcifications were classified as BI-RADS category 2, none representing recurrent disease.
CONCLUSION. Benign calcifications at the lumpectomy bed usually develop earlier than malignant calcifications, but the rate of development overlaps. Most calcifications initially placed in the probably benign category evolve quickly to more benign or more malignant morphology. Most calcifications heralding recurrence appear suspicious on first presentation.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The interval in which benign and malignant calcifications develop after treatment has been addressed by some authors, with Dershaw et al. [13] reporting benign calcifications developing earlier and Rebner et al. [10] reporting them developing later than malignant calcifications.
The purpose of this study was to evaluate the rate of development and the rate of change between benign and malignant calcifications that develop at the lumpectomy bed after breast conservation therapy to determine if these factors can aid in the mammographic management of these patients.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Patient age and date of lumpectomy were obtained by reviewing pathology records and the surgical database in 43 patients, and through review of standardized patient questionnaires in eight patients in whom pathology records were not available. Patients were 28-81 years old (mean, 55 years). Breast conservation therapy was performed from August 1982 to February 1996; only three (6%) of 51 patients had surgery before 1985. Forty-three (84%) of 51 had known radiation therapy, two had none, and radiation therapy was unknown in six. Radiation therapy doses were not readily obtainable in our patient population (many of whom returned to local centers for treatment after diagnosis at our tertiary care center), and therefore that was not correlated with development of calcifications at the lumpectomy bed.
During the period of the study, no standard mammographic surveillance protocol existed at our institution. On the basis of clinician preference, 24 patients underwent 6-month surveillance of the treated breast for 1-3 years, with yearly surveillance of the untreated breast; and 27 received yearly surveillance of both breasts after lumpectomy. Radiologists recommended close mammographic surveillance (defined as two 6-month follow-up intervals followed by yearly surveillance thereafter) for findings thought to be probably benign. Serial mammograms and reports were retrospectively reviewed to determine the time of development of calcifications after surgery. morphology, rate of change, and number (classified as 1-5, 6-10, 11-15, or >15). In one patient, only the report of negative mammographic findings was available, with the next mammogram showing new microcalcifications; this patient was included in the study. All mammograms were the film-screen type (no xero-mammograms). Spot magnification images were available for all cases considered probably benign or suspicious but were not always performed for cases considered obviously benign.
Retrospective mammographic review was performed as follows: All cases with benign findings on the initial interpretation were reviewed by one radiologist. All cases but one considered probably benign or suspicious were reviewed in consensus. The one exception was considered probably benign on initial assessment of new calcifications; it was reviewed by only one radiologist because the patient removed her images before consensus review. Mammograms showing the initial appearance of calcifications were reviewed to determine morphology and number, classified as one to five, six to 10, 11-15, or more than 15 calcifications. The initial prospective radiologic interpretation was retrospectively classified using the American College of Radiology's Breast Imaging Reporting and Data System (BI-RADS) categories [14] because BI-RADS categories were not prospectively assigned during the study period. On the basis of the follow-up recommendation made after initial detection and interpretation of new calcifications at the lumpectomy bed, BI-RADS category 2 or 3 was assigned. If a biopsy was recommended, BI-RADS category 4 or 5 was assigned. Categories 4 and 5 could not be retrospectively distinguished because the degree of prospective suspicion was not usually apparent from the initial radiology report. The prospective interpretation was preferentially used rather than a retrospective impression to avoid bias and to determine the outcome of actual mammographic practice.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Twenty-nine (55%) of 53 calcifications were initially interpreted as BI-RADS category 2. On review, all these cases had clearly coarse, dystrophic morphology. Twenty-eight were disease-free at a median follow-up of 92 months (range, 24-186 months). One developed a local recurrence elsewhere in the lumpectomy quadrant, manifesting as a spiculated mass on imaging.
Fifteen (28%) of 53 calcifications were initially interpreted as BI-RADS category 3. Morphology in these 15 cases is detailed in Table 2. Twelve (80%) of 15 were downgraded to BI-RADS category 2 at a median follow-up of 6.5 months (range, 6-16 months) on the basis of stability (n = 2) or increasingly coarse morphology (n = 10) (Fig. 1A,1B). The median disease-free interval after lumpectomy in these patients was 77 months (range, 25-146 months). Three (20%) of 15 were upgraded to BI-RADS category 4, one despite stability at the 6-month follow-up (benign at biopsy) and two for increasing pleomorphic calcifications at 6 months (malignant, local recurrence only) (Fig. 2A,2B). The case upgraded to BI-RADS category 4 at the 6-month follow-up despite stability consisted of 11-15 micro- and macrocalcifications; fat necrosis was diagnosed at biopsy. One case of local recurrence initially consisted morphologically of one to five round microcalcifications with one linear form; the number of microcalcifications increased to six to 10 with increasing pleomorphism by the 6-month follow-up. Biopsy yielded a 3-mm focus of comedo-type ductal carcinoma in situ. The second case of local recurrence (Fig. 2A,2B) initially consisted of six to 10 loosely grouped faint microcalcifications seen on only one image; the microcalcifications increased to 11-15 seen on two orthogonal images at the 6-month follow-up. Biopsy of this case yielded a 24-mm mixed infiltrating and in situ ductal local recurrence.
|
|
|
|
|
Nine (17%) of 53 calcifications were initially interpreted as BI-RADS category 4 or 5, all with suspicious morphology and distribution (Table 2). Six (67%) of nine were malignant (Fig. 3) and three (33%) of nine were benign at biopsy.
|
On the basis of the number of calcifications at initial presentation, a benign diagnosis was present in 35 (95%) of 37 cases with one to five calcifications, seven (88%) of eight cases with six to 10 calcifications, two (100%) of two cases with 11-15 calcifications, and one (17%) of six cases with more than 15 calcifications.
On the basis of morphology, no cases with coarse, dystrophic calcifications represented disease recurrence. All calcifications representing local tumor recurrence showed pleomorphism and clustered, linear or segmental distribution on either initial presentation (n = 6) or short-interval follow-up (n = 2). Four (33%) of 12 cases with pleomorphic microcalcifications represented benign disease at biopsy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Several authors have addressed the time of development of calcifications at the lumpectomy bed with respect to benign or malignant histopathology. Dershaw et al. [13] found that calcifications due to benign change developed earlier than calcifications due to malignancy; however, Rebner et al. [10] reported that malignant calcifications tended to occur earlier than benign calcifications. Vora et al. [15] developed a prospective management algorithm based in part on the time of development of calcifications at the lumpectomy bed. In that study, patients who developed microcalcifications more than 3 years after lumpectomy were offered biopsy versus close follow-up for probably benign morphology. Patients with probably benign morphology developing within 3 years of lumpectomy underwent follow-up surveillance. The algorithm of Vora et al. was based on the assumption that benign calcifications typically develop earlier after lumpectomy than malignant calcifications. Our results agree with those of Dershaw et al. [13], with the median time to development of benign calcifications (23 months) being shorter than that for malignant calcifications (39 months). However, significant overlap is seen in the time to development. Thus, although time to development may be helpful, it is not a reliable predictor of histopathology in the assessment of new calcifications. Although it has been stated that benign dystrophic-type calcifications develop 2-5 years after lumpectomy [6], some patients in our series developed these types of calcifications more than 5 years after conservative treatment, with one patient developing obviously dystrophic calcifications more than 14 years after lumpectomy.
The rate of change was a useful mammographic feature in our series. Most cases undergoing short-interval follow-up evolved quickly to more malignant or more benign histopathology, usually by the first follow-up examination. Most cases representing recurrent disease appeared suspicious on first presentation, but two cases were initially interpreted as probably benign. In both cases, the first follow-up examination at 6 months showed increasingly suspicious morphology and a change in the number of calcifications, prompting biopsy. Both patients had only local disease recurrence. Although malignant calcifications have been reported to remain stable in appearance [16], we did not observe this in the small number of local recurrences in our series.
An important feature of new calcifications at the lumpectomy bed is morphology. Although both Rebner et al. [10] and Mitnick et al. [17] reported difficulty using morphology to accurately predict which calcifications represent tumor recurrence, Dershaw et al. [18] found calcifications associated with local tumor recurrence to invariably be of indeterminate or suspicious morphology. In a study by Solin et al. [11], the calcifications in six of seven cases of local recurrence with available mammograms met the usual criteria for biopsy; six additional cases were considered suspicious morphologically but were benign on biopsy. The management algorithm of Vora et al. [15] was based on morphology, number of calcifications, and time to development. In their study, no case developing before or after 3 years with probably benign morphology represented recurrent disease. In their study, most cases developing calcifications with malignant morphology more than 3 years after lumpectomy had recurrent tumor. Although an evaluation of calcification morphology was not a primary objective of our study, our results are in agreement with the results of these authors [11, 15, 18]. In our series, no cases of calcifications with benign morphology represented recurrent disease, and most cases with suspicious morphology represented local tumor recurrence.
The number of calcifications developing at the lumpectomy bed as a prognostic indicator has also been addressed as a useful mammographic feature. In one series [18], 17 of 22 local tumor recurrences associated with calcifications had more than 10 calcifications on mammography. Vora et al. [15] used number of calcifications to decide on biopsy; if four or more calcifications developed more than 3 years after conservative therapy, biopsy was performed regardless of morphology. In our series, most cases of local recurrence manifested with more than 15 calcifications on initial presentation, whereas most cases of benign disease manifested with fewer than 10 calcifications. Logically, the more calcifications, the more easily their overall appearance can be characterized; our data suggest that most malignant calcifications can be quickly and accurately characterized as suspicious on initial detection.
The presence or absence of microcalcifications in the patient's original tumor should not influence interpretation of new microcalcifications after treatment. Although Philpotts et al. [19] reported a similar mammographic appearance for primary and locally recurrent tumor most of the time, Giess et al. [20] found that similarity between primary and locally recurrent tumor varied by histopathology. In their study, primary and locally recurrent ductal carcinoma in situ shared similar imaging features (manifested by microcalcifications), but primary invasive cancer shared similar imaging features with invasive local recurrence only 47% of the time. Another group of researchers also found no correlation between microcalcifications in the primary tumor and the cause of new calcifications at the lumpectomy bed [11]. On the basis of these studies, we did not attempt to relate the presence or absence of calcifications in the primary tumor to the development of calcifications after lumpectomy.
During the time of our study, no standard protocol existed at our institution for follow-up mammographic surveillance after breast conservation therapy. This is a study limitation. It is possible that the number of new calcifications initially classified as probably benign may be underrepresented in our data because some of these cases may have appeared obviously benign when imaged at a yearly interval. It is also possible that some cases with a suspicious appearance may have initially been classified as probably benign had they been imaged at 6-month intervals. Nevertheless, our results reflect actual mammographic practice. Even with a standardized short-interval imaging protocol for several years after lumpectomy, some benign and malignant calcifications would develop after that interval and thus be detected on yearly surveillance. We currently image our patients after lumpectomy every 6 months for 2 years after definitive surgery, with magnification images performed at the discretion of the radiologist. In our study, 23 (51%) of 45 benign and four (50%) of eight malignant calcifications developed within a 24-month period after definitive surgery (Table 1). We believe that this surveillance schedule enables us to closely follow the development and stabilization of changes in the breast after treatment and to detect early signs of local tumor recurrence.
In conclusion, our data show that although benign calcifications tend to develop earlier than malignant ones, the time to development overlaps and thus cannot be reliably used in management decisions. Most calcifications considered probably benign evolve to more suspicious or more benign morphology on short-interval mammographic follow-up, supporting the use of this management in this challenging patient population. Most malignant calcifications developing at the lumpectomy bed after conservative treatment appear suspicious on first mammographic presentation. Finally, new suspicious calcifications at the lumpectomy bed have a high likelihood (67%) of malignancy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  I. Gunhan-Bilgen and A. Oktay Management of Microcalcifications Developing at the Lumpectomy Bed After Conservative Surgery and Radiation Therapy Am. J. Roentgenol., February 1, 2007; 188(2): 393 - 398. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
