| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Breast Imaging |
1 From the Breast Health Center, California Pacific Medical Center, 3698 California St, Suite 2F, San Francisco, CA 94118. Received October 10, 2003; revision requested December 10; revision received January 5, 2004; accepted February 2. Address correspondence to F.R.M. (e-mail: liue@sutterhealth.org).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: One hundred thirteen consecutive patients underwent vacuum-assisted 11- or 14-gauge needle stereotactic core biopsy for calcifications with malignant histologic results in core samples from 116 lesions. For each lesion, calcification was identified in at least one core at specimen radiography. Cores with and those without calcification seen on magnified specimen radiographs were separately submitted to and reported on by pathologists, who obtained additional levels in cores with calcification. All patients underwent surgical excision of the lesion area within 7 weeks. The pathologic diagnosis in core samples with and those without calcification on specimen radiographs was compared with final diagnosis at surgical excision. Fisher exact test was used for all 
2 determinations of statistical significance.
RESULTS: Cores with calcification on specimen radiographs were more likely to enable a final diagnosis of malignancy than were cores without calcification (98 [84%] vs 82 [71%] of 116; P = .02). Cores without calcification were significantly more likely to cause a diagnosis of cancer to be missed than were those with calcification on specimen radiographs (13 [11%] vs one [1%] of 116; P < .001). Underestimates of malignancy were more frequent in 14- than in 11-gauge specimens (11 [18%] of 60 vs six [10%] of 56; P = .30). Regardless of needle size, there was no significant difference in underestimation of malignancy between cores with and without radiographically evident calcification (17 [15%] vs 21 [18%] of 116; P = .60).
CONCLUSION: Specimen radiography is essential to document calcification retrieval. Cores without radiographically demonstrated calcification may fail to show a malignant lesion. Separate identification of calcium-containing cores may assist the pathologist, who can more thoroughly evaluate these cores with additional levels of section.
© RSNA, 2004
Index terms: Breast, biopsy, 00.1261 • Breast neoplasms, calcification, 00.812 • Specimens, radiography, 00.128 • Stereotaxis, 00.1267
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Specimen Radiography and Surgical Excision
For all specimens, extracted cores underwent radiography in a dedicated specimen radiography unit (Faxitron X-ray, Buffalo Grove, Ill) with use of at least x2 magnification.
From each lesion, the cores with and those without calcification on specimen radiographs were separated and were submitted in labeled formalin containers for pathologic examination. For each lesion, a final pathology report separately described findings for cores with and those without radiographic calcifications. All reports were generated on the day of the core biopsy by one of eight pathologists responsible for surgical specimens. The interpreting pathologists, each of whom had at least 5 years of experience with core biopsy samples, were provided with a copy of the specimen radiographs and the knowledge of which separated cores contained radiographic calcification. Two levels were obtained routinely for all cores, and an unspecified number of additional levels was examined in those cores that demonstrated radiographic calcification. Surgical excision of the stereotactic biopsy site was performed, after wire localization, within 7 weeks of the core biopsy (mean, 2 weeks; range, 1–7 weeks) in all patients with malignant histologic results. Surgery was performed by one of four breast surgeons, each of whom had at least 15 years of experience.
Cores and Surgical Specimen Comparisons
In 140 cases of malignant calcifications, specimen radiographs and pathology reports were available for 116 lesions in 113 patients. For 24 lesions, histologic results were not reported separately for cores with and those without calcifications. These 116 cases in 113 women, therefore, constitute the final study group (mean age, 56 years; range, 32–78 years; median, 52 years). The size of target calcifications measured on mammograms, the number of cores obtained, and the number that contained calcification on specimen radiographs were recorded by each radiologist at the time of core biopsy. Table 1 summarizes the data for these 116 procedures, of which 60 were performed with 14-gauge and 56 were performed with 11-gauge needles. For small clusters, targeting was accomplished by positioning the calcifications adjacent to the needle sample notch. In larger areas, sampling was done with a minimum of six cores, which were obtained at 12-, 2-, 4-, 6-, 8-, and 10-o’clock locations. Calcification was identified on specimen radiographs of at least one core for each case (mean, 3.6 cores; median, 3.0 cores; range, one to 10 cores).
|
A cancer was judged to have been missed when the core diagnosis was benign but ductal carcinoma in situ or invasive cancer was found in the surgical specimen.
Statistical Analysis
Statistical comparison of core and surgical specimen results was performed by using the Fisher exact test for all 2 determinations of statistical significance, and a P value of less than .05 was the limit of statistical significance (3).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Two ductal carcinoma in situ calcifications, 2 and 4 mm in maximum mammographic size, were completely removed in 11-gauge cores, and one 2-mm cluster was completely removed in a 14-gauge core. In these three cases, only benign tissue was contained in the surgical specimen. Clips were deployed to mark these biopsy sites and were used to guide the wire localizations. In two other calcifications sampled with 14-gauge needles and in one cluster sampled with an 11-gauge needle, ductal carcinoma in situ was found in the cores with and those without calcium, while only atypical ductal hyperplasia remained at surgical excision. Calcifications in these three cases were also 2–4 mm in maximum size. Core biopsy results were considered to be accurate in these cases.
Table 2 compares the frequency of accurate diagnosis, underestimated cases, and missed cancers in 11- and 14-gauge cores with and without calcification demonstrated at specimen radiography. Regardless of the needle size, cores containing calcifications on specimen radiographs were more likely to enable an accurate diagnosis than were cores without calcifications (98 [84%] of 116 vs 82 [71%] of 116; P = .02). Core specimens without radiographically evident calcifications were significantly more likely to cause a diagnosis of cancer to be missed than were those with calcifications (13 [11%] of 116 vs one [1%] of 116; P < .001).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Some have suggested that tissue providing sufficient evidence to allow accurate histologic diagnosis could be obtained when cores failed to demonstrate calcification at specimen radiography (4). This assertion is not supported by published data.
The larger core samples obtained with vacuum-assisted needles efficiently extract calcifications demonstrated on specimen radiographs when these calcifications are the target of the needle biopsy (5). This was true in 99% of 160 11-gauge vacuum-assisted core biopsies reported by Cangiarella et al (6). In that study, cores with and those without calcifications were separately submitted for pathologic examination, but their histologic results were not separately reported.
In our investigation, the lesion size and the number of cores obtained were comparable between the 11- and 14-gauge cases. That no significant difference in histologic underestimation of disease could be correlated with the larger specimen size obtained with the 11-gauge instrument was unanticipated. In a multi-institutional study that compared the underestimation of malignancy in cores diagnosed as atypical ductal hyperplasia, Jackman et al (7) found there were more underestimations (26 of 54, 48%) with the 14-gauge automated-gun samples than with 14-gauge vacuum-assisted biopsies (13 of 74, 18%). The majority (78%–84%) of these lesions manifested as calcifications of 9 mm in mean size. However, when these results were adjusted for the number of cores obtained, the underestimation rates for the two techniques were similar when fewer than 10 cores were extracted (44% vs 50%). These results might be explained by the more extensive sampling of larger lesions, which can be accomplished by multi-area targeting with the automated gun.
In our study, six to eight cores were obtained in most patients, the mean lesion size was 8–10 mm, and atypical ductal hyperplasia cases represented only 10 (26%) of the 38 underestimated cases. Five (13%) of these occurred with 14-gauge samples, a rate that is lower than the 38% (nine of 24 cases) for 14-gauge vacuum-assisted core biopsies reported by Meyer et al (8) or the 39% (11 of 28 cases) reported by Darling et al (9). The remaining five atypical ductal hyperplasia underestimates in our patients occurred with 11-gauge cores. This 13% (five of 38) underestimation rate is more comparable to the 11% (one of nine) underestimation rate reported by Meyer et al and the 19% (16 of 86) rate described by Darling et al for 11-gauge biopsy procedures with similar lesion size (mean, 9.7 mm) and number of core samples (mean, 9.0).
The majority of our underestimates (28 of 38, 74%) occurred when core samples yielded ductal carcinoma in situ but invasive disease was present at surgery. Twenty-two of these 28 cases demonstrated ductal carcinoma in situ in both cores with and those without calcifications on specimen radiographs.
Jackman et al (10) have published ductal carcinoma in situ underestimation rates from 16 institutions that compared 14-gauge and 11-gauge vacuum-assisted biopsy techniques. The ductal carcinoma in situ underestimation rate was 11% (38 of 348 lesions) for 14-gauge and 11% (69 of 605 lesions) for 11-gauge samples. In 86%–91% of cases, biopsy was performed for calcifications, and, as in their study of atypical ductal hyperplasia, the number of cores obtained (grouped as fewer than 10 cores and 10 or more cores) correlated most closely with the underestimation rate. This suggests that increasing the number of core samples is more likely to result in extracting the invasive component of an infiltrating ductal cancer, if one is present.
Calcifications within an infiltrating ductal cancer are most commonly found within an in situ component of the lesion (11). These calcifications, which represent the mammographic target for core biopsy, would direct sampling to the in situ portion rather than the invasive portion of a ductal cancer. It is reasonable to conclude, therefore, that when both in situ and invasive disease are present, the in situ component is likely to be sampled in cores containing calcification. This is more likely to occur in smaller lesions exhibiting microcalcifications, as these are less likely to represent invasive cancer at surgery. Among 304 such lesions reported by Stomper et al (12), invasive foci were present in 26% (29 of 110) of malignancies associated with microcalcifications less than 10 mm in size compared with 40% (77 of 194) of cancers with microcalcifications 11 mm or larger (P = .019).
The 20.4% underestimation rate (76 of 373 cases) described by Jackman et al (10) for ductal carcinoma in situ in large-core biopsies is comparable to the 24% underestimation rate for ductal carcinoma in situ in our patients (28 of 116 cases) for both 14- and 11-gauge biopsies. The number of cores obtained in our patients was approximately half of the mean number of cores in the Jackman et al study (seven vs 13), but the mean size of lesions we sampled was also smaller (8–10 vs 13 mm). These factors may have served to offset one another, since in the study by Jackman et al, ductal carcinoma in situ underestimation diminished as lesion size decreased. The fewer cores we obtained may have been sufficient to sample these smaller lesions with a comparable rate of underestimation.
We recognize the limitations of our study imposed both by its retrospective design and by the foreknowledge of the pathologists, which possibly introduced bias into the processing and interpretation of cores with and those without radiographic calcification. Although at least two levels of section were obtained for each core, the number of levels was not recorded and varied among pathologists. Although cores containing radiographic calcification were uniformly evaluated with additional levels, the diagnostic yield of these was not specifically described in the formal pathology report.
We conclude that specimen radiography is essential to document calcification retrieval and avoid core biopsy procedures that fail to sample a malignant lesion, and identifying cores containing calcification on specimen radiographs may prompt the pathologist to obtain additional levels of section to more thoroughly examine these cores.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Authors stated no financial relationship to disclose.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E A Rakha and I O Ellis An overview of assessment of prognostic and predictive factors in breast cancer needle core biopsy specimens J. Clin. Pathol., December 1, 2007; 60(12): 1300 - 1306. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
