- Email: [email protected]
Is surgical biopsy mandatory in case of atypical ductal hyperplasia on 11-gauge core needle biopsy? a retrospective study of 300 patients
The American Journal of Surgery (2008) 196, 339 –345
Clinical Surgery-International
Is surgical biopsy mandatory in case of atypical ductal hyperplasia on 11-gauge core needle biopsy? a retrospective study of 300 patients Caroline Forgeard, M.D., Medhi Benchaib, M.D., Ph.D., Nicole Guerin, M.D., Philippe Thiesse, M.D., Hervé Mignotte, M.D., Christelle Faure, M.D., Catherine Clement-Chassagne, M.D., Isabelle Treilleux, M.D., Ph.D.* Department of Pathology, Leon Berard Centre, 28 rue Laennec, 69373 Lyon Cedex 08, France KEYWORDS: Directional vacuumassisted biopsy; Atypical ductal hyperplasia; Underestimation rate; Follow up; Surgical excision
Abstract BACKGROUND: Atypical ductal hyperplasia (ADH) is diagnosed in 4% to 10% of directional vacuum-assisted stereotactic biopsies (DVABs) performed for microcalcifications. Since the underestimation rate varies from 7% to 36%, surgical excision is still recommended, although some authors have tried to identify a subset of patients who can be spared surgery. METHODS AND RESULTS: In this study, we analyzed a retrospective series of 300 patients with ADH on 11-gauge DVAB. The only 4 events that occurred (3%) in 135 of 184 patients (61%) who were followed may not be due to underestimation. Comparing the diagnoses on DVAB and surgical excisions for 116 patients (39%), we identified 3 subsets of patients: no underestimation (size ⬍6 mm and complete removal), low rate of 4% (ⱕ2 foci ADH in microcalcifications either ⬍6 mm with incomplete removal or ⱖ6 mm and ⬍21 mm), and high rate of 36% to 38% (⬎2 foci ADH in microcalcifications either ⬍6 mm with incomplete removal or ⱖ6 mm and ⬍21 mm, lesion size ⱖ21 mm). CONCLUSIONS: Our results suggest that strict follow-up can be a safe option for the first 2 groups of patients, but that surgical excision is mandatory for patients from the third group. © 2008 Elsevier Inc. All rights reserved.
With the advent of both breast screening programs and the directional vacuum-assisted biopsy (DVAB) technique, an increasing number of borderline lesions such as atypical ductal hyperplasia (ADH) and lobular neoplasia are being diagnosed by core needle biopsies for microcalcifications. Most studies in the literature have focused on ADH since it occurs far more frequently.1–10 They have highlighted that ADH diagnosis on DVAB is associated with a rate of
* Corresponding author. Tel.: ⫹33 44 78782839; fax: ⫹33-4-78782713 E-mail address: [email protected] Manuscript received May 15, 2007; revised manuscript July 17, 2007
0002-9610/$ - see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2007.07.038
carcinoma ranging from 11% to 35% in subsequent excisions.1–13 Because of this underestimation risk, some authors have recommended a mandatory surgical biopsy while others have discussed options between surgery and follow-up. Surgical decision-making is usually based on personal or family history of breast cancer, palpable or ultrasonographic associated mass, incomplete removal of calcifications with DVAB, or unclear mammographic follow-up.8 –10 More recently some authors have used pathological criteria to identify a subset of patients who do not benefit from surgery. They assessed the extent of ADH within ducts and demonstrated that underestimation does not occur when ADH is confined to 1 to 2 ducts on biopsy.8,9,13
340
The American Journal of Surgery, Vol 196, No 3, September 2008
The aim of the present study was to identify valuable criteria for deciding between surgical biopsy and follow-up. For that purpose, we analyzed a retrospective series of 300 patients diagnosed with ADH on DVAB at our institution between February 1999 and May 2005.
Materials and Methods A total of 300 patients with ADH as the highest grade lesion on 11-gauge DVAB were retrieved from our files. All DVABs were performed for microcalcifications by welltrained radiologists at Leon Berard Center from February 1999 to May 2005, and were obtained with a dedicated stereotactic device (MammoTest; Fischer Imaging, Denver, CO). A radiography of the core samples was performed after the biopsy, as well as a mammogram to check the accuracy of the procedure. If the mammographic lesion was completely removed, a clip was placed in the biopsy cavity. The diagnoses of ADH were made on biopsies by 2 well-trained pathologists specialized in breast pathology. A diagnosis of ADH was rendered when ductal epithelial hyperplasia exhibited cytological and architectural features of low-grade carcinoma and did not exceed 2 mm in aggregate diameter. Two subclasses were identified according to the extent of ADH within ducts and/or lobules: ADH confined to 1 or 2 foci (ⱕ2 foci ADH) versus ADH involving at least 3 foci (⬎2 foci ADH). The following important criteria for decision-making between surgery or follow-up were noted prospectively in the patients’ records: age, personal or family history of breast cancer (only first-degree relatives), associated palpable or ultrasonographic mass, extent of microcalcifications (lesion size) on mammography or incomplete removal of microcalcifications with DVAB, total number of cores, number of cores with microcalcifications, and ADH extension within ducts (ⱕ2 foci ADH or ⬎2 foci ADH). At that time, since no guidelines had been established in the institution, each
Table 1
case was discussed by a multidisciplinary team including surgeons, radiologists, and pathologists, all involved in breast pathology. Ninety-one percent of the patients underwent surgery at our institution. Preoperative needle localization and intraoperative specimen radiography were performed to ensure accurate excision. Histological diagnoses at surgical excision were classified as benign or malignant. Any residual ADH or other benign lesion such as non-atypical ductal hyperplasia was classified as benign, whereas any ductal carcinoma in situ (DCIS) or invasive ductal carcinoma (IDC) was classified as malignant. Patients who did not undergo surgery were followed up by bilateral mammographies 6 months after DVAB and every year thereafter. These patients were categorized according to outcome. Recurrence was stated when a diagnosis of malignancy was determined on further procedures. Patients with no mammographic alteration or a benign diagnosis, including ADH on further biopsies, were classified into the “no evolution” group. The data were analyzed using chi-square and Student t tests, as well as Fisher exact test for the small sample. P values of less than .05 were considered statistically significant.
Results General population From February 1999 to May 2005, 2,214 biopsies (11gauge DVAB) were performed at Leon Berard Center and 300 cases of ADH were diagnosed (14% of total biopsies). One hundred sixteen patients (39%) had surgery and 184 (61%) were followed. Table 1 shows the clinical, radiological, and histological features of the 300 patients with ADH. Patients undergoing surgery (surgery group) and those who were followed (ab-
Clinical, radiological, and histological features of patients with ADH on DVAB performed for microcalcifications
Mean age, y (SD) Family history (%) Personal history (%) Associated echographic mass (%) Mean size mm (SD) Complete removal with DVAB (%) No. of cores (SD) No. of cores with microcalcifications (SD) >2 foci ADH (%) Total
Total
Surgery
Abstention
P value
55 8 21 17 14 180 16 6 83 300
54 3 11 9 18 50 16 6 72 116
55 5 10 8 12 130 16 6 11 184
NS* NS† NS‡ NS‡ .004* <.0001‡ NS* NS* <.0001‡
(8) (3) (7) (6) (15) (60) (4) (3) (28)
(8) (3) (9) (8) (17) (43) (3) (3) (62)
(8) (3) (5) (4) (12) (71) (4) (3) (6)
ADH ⫽ atypical ductal hyperplasia; DVAB ⫽ directional vacuum-assisted stereotactic biopsy; SD ⫽ standard deviation; NS ⫽ not significant. *Student t test. †Fisher exact test. ‡Chi-square test.
C. Forgeard et al.
Needle core biopsy in atypical ductal hyperplasia
stention group) are compared. The extent of microcalcifications was significantly higher (P ⫽ .004) in the surgery group (18 mm) compared to the abstention group (12 mm). In contrast, the incidences of both ⱕ2 foci ADH and complete removal of microcalcifications with DVAB were more frequent in the abstention group (P ⬍.0001). However, there was no statistical difference between the 2 groups for age, history of breast cancer, type of radiological lesion (associated microcalcifications and/or mass), number of core biopsies, and number of core biopsies with microcalcifications.
Surgery group The pathological diagnosis on surgical excision was benign in 75% of cases. Among the malignant cases, 16% were low-grade DCIS and 9% were intermediate- or highgrade DCIS or IDC. Table 2 compares the clinical, radiological, and histological findings on DVAB between the “benign” and “malignant” groups as defined by the final diagnosis on surgical excision. The mean size of the radiological lesion was larger in the “malignant” group than in the “benign” group, but this was significant only for those patients with ⱕ2 foci ADH (P ⫽ .024). Complete removal of microcalcifications with DVAB was more frequent in the benign group (46% vs 35%), whereas incomplete removal was more often seen in the malignant group (65% vs 54%). These differences did not reach significance. More importantly, malignancy in the
Table 2 Comparison of clinical, radiological and histological findings on DVAB between patients with a “benign” or a “malignant” diagnosis on surgical excision
Age, y (SD) Family history (%) Personal history (%) Associated echographic mass (%) Mean lesion size in mm (SD): <2 foci ADH (SD) ⬎2 foci ADH (SD) Incomplete removal with DVAB (%) No. of cores (SD) No. of cores with microcalcifications (SD) >2 foci ADH (%) Total (%)
Benign
Malignant
P value
55 (8) 2 (2) 10 (11)
53 (7) 1 (3) 1 (3)
NS* NS† NS†
9 (10)
1 (3)
NS‡
16 (14) 19 (15) 14 (14)
23 (22) 37 (26) 21 (20)
NS* .02§ NS§
47 (54) 16 (3)
19 (65) 17 (4)
NS‡ NS*
6 (3) 48 (55) 87 (75)
7 (3) 24 (83) 29 (25)
NS* .008‡
DVAB ⫽ directional vacuum-assisted stereotactic biopsy; ADH ⫽ atypical ductal hyperplasia; SD ⫽ standard deviation; NS ⫽ not significant. *Student t test. †Fisher exact test. ‡Chi-square test. §t test.
341
Table 3 Histological findings on surgical excision according to histological findings on DVAB
Benign ADH Low-grade DCIS Intermediate-grade DCIS High-grade DCIS Invasive ductal carcinoma Total
ⱕ2 foci ADH
⬎2 foci ADH
16 23 4
20 28 15
0 0
5 2
1 44
2 72
Total (%) 36 (31%) 51 (44%) 19 (16%) 5 (4%) 2 (2%) 3 (3%) 116 (100%)
DVAB ⫽ directional vacuum-assisted stereotactic biopsy; ADH ⫽ atypical ductal hyperplasia; DCIS ⫽ ductal carcinoma in situ.
surgical specimen was significantly (P ⫽ .008) more frequent with more than 2 foci ADH on DVAB than with ⱕ2 foci ADH. Table 3 summarizes the correlation between the final diagnosis on surgical excision and the extent of ADH on DVAB. Among 44 patients with ⱕ2 foci ADH, only 4 had low-grade DCIS (9%) and 1 had IDC (2%) at excision. There was no intermediate- or high-grade DCIS. Among 72 patients with more than 2 foci ADH, 15 were diagnosed with low-grade DCIS (20%), 5 with intermediate-grade DCIS (7%), 2 with high-grade DCIS (3%), and 2 with IDC (3%). Using a logistic regression analysis, the significant risk factors for malignancy were the lesion size and the extent of ADH (Table 4). The risk of malignancy was reduced both for patients with a lesion size ⬍21 mm (odds ratio .5, P ⬍.05) with a lower relative risk when the size was ⬍6 mm (odds ratio .13, P ⬍.01) in comparison to patients with lesion size ⱖ21 mm, and for patients with ⱕ2 foci ADH in comparison with more than 2 foci ADH (odds ratio .2, P ⬍.01). However, the complete removal of the lesion did not reduce the risk of malignancy significantly in comparison with incomplete removal. Table 5 combines the previous parameters in a multivariate analysis. As shown, no malignancy was observed when Table 4 Relative risk of malignancy on surgical excision using a logistic regression analysis
Complete removal with DVAB Lesion size < 6 mm 6 mm < lesion size <21 mm <2 foci ADH
Odds ratio
95% CI
P value
1.077 .128
.540–2.149 .029–.561
1.077 .006
.495 .202
.253–.968 .069–.594
.040 .004
DVAB ⫽ directional vacuum-assisted stereotactic biopsy; ADH ⫽ atypical ductal hyperplasia; CI ⫽ confidence interval.
342
The American Journal of Surgery, Vol 196, No 3, September 2008
Table 5 Multivariate analysis of histological findings on surgical excision according to complete/incomplete removal, lesion size, extension of ADH on DVAB Removal
Lesion size
ADH extension
Benign
Malignant
Total
Complete
⬍6 mm
Complete
ⱖ6 mm
Incomplete
⬍21 mm
Incomplete
ⱖ21 mm
ⱕ2 ⬎2 ⱕ2 ⬎2 ⱕ2 ⬎2 ⱕ2 ⬎2
5 12 9 14 14 12 11 10 87
0 0 1 9 0 7 4 8 29
5 12 10 23 14 19 15 18 116
foci foci foci foci foci foci foci foci
ADH ADH ADH ADH ADH ADH ADH ADH
Total
DVAB ⫽ directional vacuum-assisted stereotactic biopsy; ADH ⫽ atypical ductal hyperplasia.
both the lesion size was below 6 mm and the lesion was completely removed. When the lesion size was 21 mm or greater, the incidence of malignancy increased dramatically (12/33). In these subgroups, the extent of ADH within the ducts had no impact (⬍6 mm and completely removed) or little impact (ⱖ21 mm) on underestimation: 4/15 in the ⱕ2 foci ADH group compared to 8/18 in the group with greater than 2 foci ADH. However, when the lesion size was ⱖ6 and less than 21 mm or when the lesion size was less than 6 mm and incompletely removed, the incidence of malignancy rose from 1/24 with ⱕ2 foci ADH (4%) to 16/42 with more than 2 foci ADH (38%).
Abstention group The follow-up was not available for 49/184 patients. The mean follow-up time for the remaining 135 patients was 30 months (range 6 months to 6 years). Four patients recurred (3%) within 1 to 5 years after initial diagnosis of ADH on DVAB. Table 6 reports the clinical and histological characteristics of these 4 tumors.
Table 6 Clinical and histological characteristics of recurrences in followed patients Lesion size
Initial biopsy
Time to recurrence
Unknown
ⱕ2 foci ADH
5 years
13 mm
ⱕ2 foci ADH
5 years
30 mm 4 mm
⬎2 foci ADH ⬎2 foci ADH
3 years 1 year
Surgical excision DCIS intermediate grade DCIS intermediate grade DCIS low grade Invasive ductal carcinoma
DCIS ⫽ ductal carcinoma in situ; ADH ⫽ atypical ductal hyperplasia.
Comments Various rates of ADH in DVAB are reported in the literature, ranging from 4% to 10%.2,4,6,8 –10,12,14 The overall ADH rate in our series was 14%, which is much higher than reported elsewhere. The difference might be due to difficulties in histological interpretation, patient selection bias, lesion size, and/or differences in needle-gauge as further described. Table 7 compares our findings with the largest published series of ADH on DVAB with regards to the number of procedures, the rate of ADH, the lesion size, the number of cores, the diameter of the needle (11/14 gauge), and the underestimation rates function of the final histological findings on surgical excision.2– 4,6,8 –10,12–14 Only 3 reports have specified the size of the lesion.8,10,12 In contrast to our series, both 11- and 14-gauge biopsies have been performed, which makes comparison difficult.8,10,12 Nevertheless, the mean size of the lesion in the current study seems to be almost twice as large as in other studies. This factor, together with the larger number of core biopsies (16 vs 11.5) and the larger diameter of the needle (11 gauge only) might explain the higher rate of ADH reported in our study. One could assume that the likelihood of ADH on DVAB might increase with more sampling, especially when larger lesions are removed. Only 2 of the other 9 studies have presented the follow-up of non-operated patients with evidence of ADH on DVAB. Both were conducted in France, where surgical excision after ADH on DVAB is not the rule. Indeed, 65% of the patients in these 2 series were spared surgery, which is similar to our population. For Plantade et al,4 the existence of one of the following criteria led to surgery: personal or family history of breast cancer, lesion size greater than 10 mm, presence of ADH on the last cores removed, or concomitant atypical lobular hyperplasia on DVAB. However, this study did not compare the outcomes of patients who underwent surgery and those who did not. In the Travade et al14 series, only patients with complete removal of the lesion by DVAB or those above 70 years of age who had no other risk factors did not undergo surgery. In our series, although personal or family history of breast cancer
(7) (7) (7) (4) (4) (6) (10) (6) (14)
ADH ⫽ atypical ductal hyperplasia; DVAB ⫽ directional vacuum-assisted stereotactic biopsy; IDC ⫽ infiltrative ductal carcinoma; DCIS ⫽ ductal carcinoma in situ; G1 ⫽ low grade; G2 ⫽ intermediate grade; G3 ⫽ high grade.
— — 0 (0) — 2 (2) — — — — — — 2 (2) — — 4 (9) — 8 (8) — — — — — — 5 (4) — — 11 (24) — 9 (8) — — — — — — 19 (16) (11) (4) (33) (14) (18) (5) (9) (13) (19) (19) (14) (22) 7 2 15 13 19 2 6 11 7 6 88 26 (3) (9) (3) (0) (3) (2) (8) (6) (8) (0) (4) (3) 2 4 2 0 3 1 5 5 3 0 25 3 (14) (13) (36) (14) (21) (7) (17) (19) (27) (19) (18) (25) 9/62 6/46 17/46 13/95 22/104 3/42 11/65 16/86 10/37 6/31 113/614 29/116 11 11 — 11/14 11 11/14 11 11/14 11 11 — 11 6 12 12 5 18 11 15 7 18 — 12 16 — — — 1–7 mm (2) — 3–85 mm (16) — 2–41 mm (3) — — 1–85 mm (9) 2–90 mm (18) (8) (5)
90 46 — 216 131 61 77 148 135 62 966 300 1,081 851 — 3,026 1,964 824 1,750 3,873 2,130 633 16,132 2,214
Mean no. of cores Lesion size (mean) No. of ADH (%) No. of DAVB
Needle core biopsy in atypical ductal hyperplasia
Adrales 20009 Burak 20002 Ely 200113 Renshaw 200110 Jackman 20023 Sneige 20038 Winchester 20036 Darling 200012 Plantade 20044 Travade 200614 Total Present study
Table 7
Underestimation of ADH at stereotactic DVAB: findings in the literature
Needle gauge
No. of underestimated/ surgery (%)
IDC (%)
DCIS (%)
DCIS G1 (%)
DCIS G2 (%)
DCIS G3 (%)
C. Forgeard et al.
343
and the type of radiological abnormality (microcalcifications associated or not with echographic mass) influenced the choice of abstention versus surgery, these data were not statistically different between the 2 groups of patients. Only the following criteria were statistically different between the 2 groups: lower average size of the lesion in the abstention group (12 vs 18, P ⬍.0001), more frequent complete removal of microcalcifications with DVAB (71% vs 43%, P ⬍.0001), and ⱕ2 foci ADH on biopsy (94% vs 38%, P ⬍.0001). Follow-up was performed by mammography at 3 months14 or at 6 months4,14 after DVAB, as was the case in our series. Plantade et al4 recommended further follow-up at 6 months and every year thereafter, whereas Travade et al14 proposed yearly mammography. The follow-up durations were 29 months (Plantade et al4) and 35.5 months (Travade et al14) in previous studies, and 30 months in ours. Interestingly, no event was noted in the series with the longest follow-up, whereas 2 invasive nonpleomorphic lobular carcinomas were observed in the study by Plantade et al4 (a condition not usually associated with microcalcifications). One can assume that ADH was a risk marker only in the vicinity of the tumor. Four events occurred in our series at, respectively, 5 years (1 intermediategrade DCIS), 3 years (1 low-grade and 1 intermediate-grade DCIS), and 1 year (one 8-mm Scarff-Bloom-Richardson grade 1 IDC) after the procedure (Table 6). All in situ carcinomas developed at least 3 years after ADH diagnosis, which may not indicate underestimation of malignancy by DVAB. The results might be related to ADH as a risk marker of breast cancer. The invasive carcinoma also may not be due to underestimation even though it happened within 1 year, because it occurred elsewhere in the breast and was suspected because of new microcalcifications. The initial lesion was only 4 mm and microcalcifications were completely removed by DVAB. Taken together, the data indicate that strict follow-up of ADH seems a safe alternative to surgery under certain circumstances (lesion ⬍10 mm, complete removal of microcalcifications by DVAB, and ⱕ2 foci ADH on DVAB). In contrast to “follow-up studies,” many authors have assessed the correlations between histological diagnoses on biopsies and surgical specimens. All have pointed out the underestimation risk. Initial studies using different needle diameters have shown underestimation rates between 11% and 67%. All have concluded that surgery must be performed in cases of ADH.1,3,5,6,11 More recent studies using DVAB and larger needles have reported lower underestimation rates of 0% to 38%.1,2,7,11,12 Table 7 shows the underestimation rates in the main series of the literature (10 recent series with more than 40 cases each).2– 4,6,8 –10,12–14 In our study, the underestimation rate for IDC was 3%, which is quite similar to average values (4%) from the main series in the literature. For DCIS, our underestimation rate was 22% compared to an average of 14% obtained from pooling the results of the main series. This discrepancy
344
The American Journal of Surgery, Vol 196, No 3, September 2008
might be due to low reproducibility in the diagnosis of intraductal proliferative lesions and/or to a patient selection bias. Nevertheless, the underestimation rate in the main series varied from 4% to 33% and our result is within this range. The grade of DCIS was specified only in 2 of the 8 series.3,13 The proportion of IDC is quite similar in the different series. In contrast, the distribution of DCIS grades seems to vary widely from one series to another, which is possibly due to the variable definition of DCIS in the literature. Either the diagnosis is based only on nuclear size or it also takes into account the presence or the absence of comedo necrosis, thus leading to poor diagnostic reproducibility.15–18 In all series, low-grade DCIS is the more frequent (8%–24%), followed by intermediate-grade DCIS (4%–9%) and high-grade DCIS (0%–2%). Since ADH and low-grade DCIS are morphologically very similar and distinguished only by the extension of the lesion, one can expect that low-grade DCIS represents the most frequent diagnosis (⬃60%) where underestimation occurs. Predictors of ADH underestimation include personal history of breast cancer,3,9 presence of palpable or ultrasonographic mass,11,19 large lesion size,3,10 incomplete removal of lesion with DVAB,3,4,8 –11,14,20 low number of cores,4,21,22 and ADH within at least 3 foci.8,9,11,13 In a large series of DVAB (N ⫽ 104), Jackman et al3 showed an important decrease in underestimation rate when patients had no personal history of cancer, a lesion size below 10 mm, and complete removal of the lesion. When these 3 criteria were met, the underestimation rate remained above 2% and the authors concluded that surgery was necessary. Renshaw et al10 in a large series (N ⫽ 95), but also Travade et al14 (N ⫽ 31) and Liberman et al20 (N ⫽ 10) in smaller studies, reported no underestimation when the lesion was totally removed by DVAB. In our series, the size of the lesion was higher in underestimated patients than in others (23 mm vs 16 mm). Moreover, when the lesion was ⱖ21 mm, the underestimation rate was high (36%), suggesting that it was probably due to unavoidable inadequate sampling of needle biopsies. In contrast, the relative risk of malignancy with a lesion size less than 21 mm was reduced (odds ratio .5, P ⬍.05), with an even lower risk in the group with lesions less than 6 mm (odds ratio .13, P ⬍.01). Indeed, we were able to demonstrate that below 6 mm, no malignancy was diagnosed on the surgical specimen when the lesion was completely removed by DVAB. Plantade et al4 found a significant decrease in the underestimation rate when the number of cores was at least 15. In our study, since the number of core biopsies was always high (average of 16), this factor could not come out. More recently several authors have described histological factors such as the extent of ADH on DVAB that would predict underestimation.8,9,13 Indeed, this particular factor was as a major criterion in our institution for choosing between surgery and follow-up: the proportion of patients with greater than 2 foci ADH was 28% in the total population compared to 62% in
the surgical group. In a series of 62 patients, Adrales et al9 also distinguished 2 populations of ADH with “mild or moderate atypia.” The risk of underestimation was approximately 9% in the first group and 44% in the second. However, these authors gave no definition of their diagnoses. Since underestimation was not observed in patients with both no personal or family history of breast cancer and mild atypia, they did not recommend surgical excision. Ely et al13 in a series of 46 patients distinguished 3 groups according to ADH extension: when ADH was confined to 2 foci or less no underestimation was seen. When ADH involved 3 foci or more than 4 foci, the underestimation rates were 50% and 87%, respectively. Since the underestimation rate increased proportionally with ADH extension, the authors concluded that when ADH is confined to 2 foci or less, surgical excision can be avoided. Sneige et al8 in a series of 42 patients found no underdiagnosis when ADH was confined to less than 3 foci and/or when the lesion was completely removed by DVAB. To our knowledge our series is the largest. The underestimation rate was 25% when the diagnosis of ADH without specification was made on DVAB. At our institution, since 1999 the extension of ADH within ducts on DVAB has been prospectively used as histological marker. In contrast to Ely et al13 our categories included only 2 levels: ⱕ2 foci ADH or greater than 2 foci ADH. Using these criteria, the underestimation rate in operated patients fell from 33% (⬎2 foci ADH) to 11% (ⱕ2 foci ADH). Moreover, in multivariate analysis the extent of ADH was the most important criterion when the size of the lesion was ⱖ6 mm and less than 21 mm and also with a lesion size less than 6 mm with incomplete removal with DVAB. In both cases, underestimation rates rose from 4% with ⱕ2 foci ADH to 38% with greater than 2 foci ADH. However, the extension of ADH on DVAB was not discriminating when the size of the lesion was less than 6 mm and completely removed (no underestimation) or when the size was ⱖ21 mm (36% total underestimation rate) with rates rising from 27% to 45% when ⱕ2 foci ADH or greater than 2 foci ADH were diagnosed, respectively.
Figure 1 Proposition of management for patients with ADH on directional vacuum-assisted stereotactic biopsy.
C. Forgeard et al.
Needle core biopsy in atypical ductal hyperplasia
Conclusion Although surgical excision is usually recommended for ADH on 11-gauge DVAB, recent data and results from this series tend to indicate that strict follow-up can be a safe alternative in a subset of patients representing almost 60% of the patients in the main series. In the present study, we have defined 3 criteria that can help in choosing between “surgery” or “follow-up.” These criteria include lesion size, complete removal of microcalcifications with DVAB, and extent of ADH. From these data, we deduced a schema of management for patients with ADH on DVAB as seen in Figure 1. In contrast to ⱖ21-mm lesions (right), surgery may be spared in cases with microcalcifications completely removed less than 6 mm (left). However, when the lesion size is less than 6 mm but with incomplete removal or when the size is ⱖ6 mm and less than 21 mm, abstention can be recommended for ⱕ2 foci ADH patients, whereas surgery seems mandatory in cases of greater than 2 foci ADH. We are currently using these new guidelines prospectively at our institution and will assess them in the future.
References 1. Zhao L, Freimanis R, Bergman S, et al. Biopsy needle technique and the accuracy of diagnosis of atypical ductal hyperplasia for mammographic abnormalities. Am Surg 2003;69:757– 62. 2. Burak WE Jr, Owens KE, Tighe MB, et al. Vacuum-assisted stereotactic breast biopsy: histologic underestimation of malignant lesions. Arch Surg 2000;135:700 –3. 3. Jackman RJ, Birdwell RL, Ikeda DM. Atypical ductal hyperplasia: can some lesions be defined as probably benign after stereotactic 11-gauge vacuum-assisted biopsy, eliminating the recommendation for surgical excision? Radiology 2002;224:548 –54. 4. Plantade R, Hammou JC, Fighiera M, et al. [Underestimation of breast carcinoma with 11-gauge stereotactically guided directional vacuumassisted biopsy]. J Radiol 2004;85:391– 401. 5. Cangiarella J, Waisman J, Symmans WF, et al. Mammotome core biopsy for mammary microcalcification: analysis of 160 biopsies from 142 women with surgical and radiologic followup. Cancer 2001;91:173–7. 6. Winchester DJ, Bernstein JR, Jeske JM, et al. Upstaging of atypical ductal hyperplasia after vacuum-assisted 11-gauge stereotactic core needle biopsy. Arch Surg 2003;138:619 –22. 7. Burbank F. Stereotactic breast biopsy of atypical ductal hyperplasia and ductal carcinoma in situ lesions: improved accuracy with directional, vacuum-assisted biopsy. Radiology 1997;202:843–7.
345
8. Sneige N, Lim SC, Whitman GJ, et al. Atypical ductal hyperplasia diagnosis by directional vacuum-assisted stereotactic biopsy of breast microcalcifications. Considerations for surgical excision. Am J Clin Pathol 2003;119:248 –53. 9. Adrales G, Turk P, Wallace T, et al. Is surgical excision necessary for atypical ductal hyperplasia of the breast diagnosed by Mammotome? Am J Surg 2000;180:313–5. 10. Renshaw AA, Cartagena N, Schenkman RH, et al. Atypical ductal hyperplasia in breast core needle biopsies. Correlation of size of the lesion, complete removal of the lesion, and the incidence of carcinoma in follow-up biopsies. Am J Clin Pathol 2001;116:92– 6. 11. Jacobs TW, Connolly JL, Schnitt SJ. Nonmalignant lesions in breast core needle biopsies: to excise or not to excise? Am J Surg Pathol 2002;26:1095–110. 12. Darling ML, Smith DN, Lester SC, et al. Atypical ductal hyperplasia and ductal carcinoma in situ as revealed by large-core needle breast biopsy: results of surgical excision. AJR Am J Roentgenol 2000;175: 1341– 6. 13. Ely KA, Carter BA, Jensen RA, et al. Core biopsy of the breast with atypical ductal hyperplasia: a probabilistic approach to reporting. Am J Surg Pathol 2001;25:1017–21. 14. Travade A, Isnard A, Bouchet F, et al. [Non-palpable breast lesions and core needle biopsy with Mammotome 11G: is surgery required in patients with atypical ductal hyperplasia?]. J Radiol 2006;87:307–10. 15. Tavassoli FA. Ductal carcinoma in situ: introduction of the concept of ductal intraepithelial neoplasia. Mod Pathol 1998;11:140 –54. 16. Page DL, Dupont WD, Rogers LW, et al. Atypical hyperplastic lesions of the female breast. A long-term follow-up study. Cancer 1985;55: 2698 –708. 17. Ghofrani M, Tapia B, Tavassoli FA. Discrepancies in the diagnosis of intraductal proliferative lesions of the breast and its management implications: results of a multinational survey. Virchows Arch 2006; 449:609 –16. 18. Lennington WJ, Jensen RA, Dalton LW, et al. Ductal carcinoma in situ of the breast. Heterogeneity of individual lesions. Cancer 1994;73: 118 –24. 19. Liberman L, Dershaw DD, Glassman JR, et al. Analysis of cancers not diagnosed at stereotactic core breast biopsy. Radiology 1997;203: 151–7. 20. Liberman L, Smolkin JH, Dershaw DD, et al. Calcification retrieval at stereotactic, 11-gauge, directional, vacuum-assisted breast biopsy. Radiology 1998;208:251– 60. 21. Jackman RJ, Burbank F, Parker SH, et al. Atypical ductal hyperplasia diagnosed at stereotactic breast biopsy: improved reliability with 14gauge, directional, vacuum-assisted biopsy. Radiology 1997;204: 485– 8. 22. Jackman RJ, Nowels KW, Shepard MJ, et al. Stereotaxic large-core needle biopsy of 450 nonpalpable breast lesions with surgical correlation in lesions with cancer or atypical hyperplasia. Radiology 1994; 193:91–5.
Clinical Surgery-International
Is surgical biopsy mandatory in case of atypical ductal hyperplasia on 11-gauge core needle biopsy? a retrospective study of 300 patients Caroline Forgeard, M.D., Medhi Benchaib, M.D., Ph.D., Nicole Guerin, M.D., Philippe Thiesse, M.D., Hervé Mignotte, M.D., Christelle Faure, M.D., Catherine Clement-Chassagne, M.D., Isabelle Treilleux, M.D., Ph.D.* Department of Pathology, Leon Berard Centre, 28 rue Laennec, 69373 Lyon Cedex 08, France KEYWORDS: Directional vacuumassisted biopsy; Atypical ductal hyperplasia; Underestimation rate; Follow up; Surgical excision
Abstract BACKGROUND: Atypical ductal hyperplasia (ADH) is diagnosed in 4% to 10% of directional vacuum-assisted stereotactic biopsies (DVABs) performed for microcalcifications. Since the underestimation rate varies from 7% to 36%, surgical excision is still recommended, although some authors have tried to identify a subset of patients who can be spared surgery. METHODS AND RESULTS: In this study, we analyzed a retrospective series of 300 patients with ADH on 11-gauge DVAB. The only 4 events that occurred (3%) in 135 of 184 patients (61%) who were followed may not be due to underestimation. Comparing the diagnoses on DVAB and surgical excisions for 116 patients (39%), we identified 3 subsets of patients: no underestimation (size ⬍6 mm and complete removal), low rate of 4% (ⱕ2 foci ADH in microcalcifications either ⬍6 mm with incomplete removal or ⱖ6 mm and ⬍21 mm), and high rate of 36% to 38% (⬎2 foci ADH in microcalcifications either ⬍6 mm with incomplete removal or ⱖ6 mm and ⬍21 mm, lesion size ⱖ21 mm). CONCLUSIONS: Our results suggest that strict follow-up can be a safe option for the first 2 groups of patients, but that surgical excision is mandatory for patients from the third group. © 2008 Elsevier Inc. All rights reserved.
With the advent of both breast screening programs and the directional vacuum-assisted biopsy (DVAB) technique, an increasing number of borderline lesions such as atypical ductal hyperplasia (ADH) and lobular neoplasia are being diagnosed by core needle biopsies for microcalcifications. Most studies in the literature have focused on ADH since it occurs far more frequently.1–10 They have highlighted that ADH diagnosis on DVAB is associated with a rate of
* Corresponding author. Tel.: ⫹33 44 78782839; fax: ⫹33-4-78782713 E-mail address: [email protected] Manuscript received May 15, 2007; revised manuscript July 17, 2007
0002-9610/$ - see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2007.07.038
carcinoma ranging from 11% to 35% in subsequent excisions.1–13 Because of this underestimation risk, some authors have recommended a mandatory surgical biopsy while others have discussed options between surgery and follow-up. Surgical decision-making is usually based on personal or family history of breast cancer, palpable or ultrasonographic associated mass, incomplete removal of calcifications with DVAB, or unclear mammographic follow-up.8 –10 More recently some authors have used pathological criteria to identify a subset of patients who do not benefit from surgery. They assessed the extent of ADH within ducts and demonstrated that underestimation does not occur when ADH is confined to 1 to 2 ducts on biopsy.8,9,13
340
The American Journal of Surgery, Vol 196, No 3, September 2008
The aim of the present study was to identify valuable criteria for deciding between surgical biopsy and follow-up. For that purpose, we analyzed a retrospective series of 300 patients diagnosed with ADH on DVAB at our institution between February 1999 and May 2005.
Materials and Methods A total of 300 patients with ADH as the highest grade lesion on 11-gauge DVAB were retrieved from our files. All DVABs were performed for microcalcifications by welltrained radiologists at Leon Berard Center from February 1999 to May 2005, and were obtained with a dedicated stereotactic device (MammoTest; Fischer Imaging, Denver, CO). A radiography of the core samples was performed after the biopsy, as well as a mammogram to check the accuracy of the procedure. If the mammographic lesion was completely removed, a clip was placed in the biopsy cavity. The diagnoses of ADH were made on biopsies by 2 well-trained pathologists specialized in breast pathology. A diagnosis of ADH was rendered when ductal epithelial hyperplasia exhibited cytological and architectural features of low-grade carcinoma and did not exceed 2 mm in aggregate diameter. Two subclasses were identified according to the extent of ADH within ducts and/or lobules: ADH confined to 1 or 2 foci (ⱕ2 foci ADH) versus ADH involving at least 3 foci (⬎2 foci ADH). The following important criteria for decision-making between surgery or follow-up were noted prospectively in the patients’ records: age, personal or family history of breast cancer (only first-degree relatives), associated palpable or ultrasonographic mass, extent of microcalcifications (lesion size) on mammography or incomplete removal of microcalcifications with DVAB, total number of cores, number of cores with microcalcifications, and ADH extension within ducts (ⱕ2 foci ADH or ⬎2 foci ADH). At that time, since no guidelines had been established in the institution, each
Table 1
case was discussed by a multidisciplinary team including surgeons, radiologists, and pathologists, all involved in breast pathology. Ninety-one percent of the patients underwent surgery at our institution. Preoperative needle localization and intraoperative specimen radiography were performed to ensure accurate excision. Histological diagnoses at surgical excision were classified as benign or malignant. Any residual ADH or other benign lesion such as non-atypical ductal hyperplasia was classified as benign, whereas any ductal carcinoma in situ (DCIS) or invasive ductal carcinoma (IDC) was classified as malignant. Patients who did not undergo surgery were followed up by bilateral mammographies 6 months after DVAB and every year thereafter. These patients were categorized according to outcome. Recurrence was stated when a diagnosis of malignancy was determined on further procedures. Patients with no mammographic alteration or a benign diagnosis, including ADH on further biopsies, were classified into the “no evolution” group. The data were analyzed using chi-square and Student t tests, as well as Fisher exact test for the small sample. P values of less than .05 were considered statistically significant.
Results General population From February 1999 to May 2005, 2,214 biopsies (11gauge DVAB) were performed at Leon Berard Center and 300 cases of ADH were diagnosed (14% of total biopsies). One hundred sixteen patients (39%) had surgery and 184 (61%) were followed. Table 1 shows the clinical, radiological, and histological features of the 300 patients with ADH. Patients undergoing surgery (surgery group) and those who were followed (ab-
Clinical, radiological, and histological features of patients with ADH on DVAB performed for microcalcifications
Mean age, y (SD) Family history (%) Personal history (%) Associated echographic mass (%) Mean size mm (SD) Complete removal with DVAB (%) No. of cores (SD) No. of cores with microcalcifications (SD) >2 foci ADH (%) Total
Total
Surgery
Abstention
P value
55 8 21 17 14 180 16 6 83 300
54 3 11 9 18 50 16 6 72 116
55 5 10 8 12 130 16 6 11 184
NS* NS† NS‡ NS‡ .004* <.0001‡ NS* NS* <.0001‡
(8) (3) (7) (6) (15) (60) (4) (3) (28)
(8) (3) (9) (8) (17) (43) (3) (3) (62)
(8) (3) (5) (4) (12) (71) (4) (3) (6)
ADH ⫽ atypical ductal hyperplasia; DVAB ⫽ directional vacuum-assisted stereotactic biopsy; SD ⫽ standard deviation; NS ⫽ not significant. *Student t test. †Fisher exact test. ‡Chi-square test.
C. Forgeard et al.
Needle core biopsy in atypical ductal hyperplasia
stention group) are compared. The extent of microcalcifications was significantly higher (P ⫽ .004) in the surgery group (18 mm) compared to the abstention group (12 mm). In contrast, the incidences of both ⱕ2 foci ADH and complete removal of microcalcifications with DVAB were more frequent in the abstention group (P ⬍.0001). However, there was no statistical difference between the 2 groups for age, history of breast cancer, type of radiological lesion (associated microcalcifications and/or mass), number of core biopsies, and number of core biopsies with microcalcifications.
Surgery group The pathological diagnosis on surgical excision was benign in 75% of cases. Among the malignant cases, 16% were low-grade DCIS and 9% were intermediate- or highgrade DCIS or IDC. Table 2 compares the clinical, radiological, and histological findings on DVAB between the “benign” and “malignant” groups as defined by the final diagnosis on surgical excision. The mean size of the radiological lesion was larger in the “malignant” group than in the “benign” group, but this was significant only for those patients with ⱕ2 foci ADH (P ⫽ .024). Complete removal of microcalcifications with DVAB was more frequent in the benign group (46% vs 35%), whereas incomplete removal was more often seen in the malignant group (65% vs 54%). These differences did not reach significance. More importantly, malignancy in the
Table 2 Comparison of clinical, radiological and histological findings on DVAB between patients with a “benign” or a “malignant” diagnosis on surgical excision
Age, y (SD) Family history (%) Personal history (%) Associated echographic mass (%) Mean lesion size in mm (SD): <2 foci ADH (SD) ⬎2 foci ADH (SD) Incomplete removal with DVAB (%) No. of cores (SD) No. of cores with microcalcifications (SD) >2 foci ADH (%) Total (%)
Benign
Malignant
P value
55 (8) 2 (2) 10 (11)
53 (7) 1 (3) 1 (3)
NS* NS† NS†
9 (10)
1 (3)
NS‡
16 (14) 19 (15) 14 (14)
23 (22) 37 (26) 21 (20)
NS* .02§ NS§
47 (54) 16 (3)
19 (65) 17 (4)
NS‡ NS*
6 (3) 48 (55) 87 (75)
7 (3) 24 (83) 29 (25)
NS* .008‡
DVAB ⫽ directional vacuum-assisted stereotactic biopsy; ADH ⫽ atypical ductal hyperplasia; SD ⫽ standard deviation; NS ⫽ not significant. *Student t test. †Fisher exact test. ‡Chi-square test. §t test.
341
Table 3 Histological findings on surgical excision according to histological findings on DVAB
Benign ADH Low-grade DCIS Intermediate-grade DCIS High-grade DCIS Invasive ductal carcinoma Total
ⱕ2 foci ADH
⬎2 foci ADH
16 23 4
20 28 15
0 0
5 2
1 44
2 72
Total (%) 36 (31%) 51 (44%) 19 (16%) 5 (4%) 2 (2%) 3 (3%) 116 (100%)
DVAB ⫽ directional vacuum-assisted stereotactic biopsy; ADH ⫽ atypical ductal hyperplasia; DCIS ⫽ ductal carcinoma in situ.
surgical specimen was significantly (P ⫽ .008) more frequent with more than 2 foci ADH on DVAB than with ⱕ2 foci ADH. Table 3 summarizes the correlation between the final diagnosis on surgical excision and the extent of ADH on DVAB. Among 44 patients with ⱕ2 foci ADH, only 4 had low-grade DCIS (9%) and 1 had IDC (2%) at excision. There was no intermediate- or high-grade DCIS. Among 72 patients with more than 2 foci ADH, 15 were diagnosed with low-grade DCIS (20%), 5 with intermediate-grade DCIS (7%), 2 with high-grade DCIS (3%), and 2 with IDC (3%). Using a logistic regression analysis, the significant risk factors for malignancy were the lesion size and the extent of ADH (Table 4). The risk of malignancy was reduced both for patients with a lesion size ⬍21 mm (odds ratio .5, P ⬍.05) with a lower relative risk when the size was ⬍6 mm (odds ratio .13, P ⬍.01) in comparison to patients with lesion size ⱖ21 mm, and for patients with ⱕ2 foci ADH in comparison with more than 2 foci ADH (odds ratio .2, P ⬍.01). However, the complete removal of the lesion did not reduce the risk of malignancy significantly in comparison with incomplete removal. Table 5 combines the previous parameters in a multivariate analysis. As shown, no malignancy was observed when Table 4 Relative risk of malignancy on surgical excision using a logistic regression analysis
Complete removal with DVAB Lesion size < 6 mm 6 mm < lesion size <21 mm <2 foci ADH
Odds ratio
95% CI
P value
1.077 .128
.540–2.149 .029–.561
1.077 .006
.495 .202
.253–.968 .069–.594
.040 .004
DVAB ⫽ directional vacuum-assisted stereotactic biopsy; ADH ⫽ atypical ductal hyperplasia; CI ⫽ confidence interval.
342
The American Journal of Surgery, Vol 196, No 3, September 2008
Table 5 Multivariate analysis of histological findings on surgical excision according to complete/incomplete removal, lesion size, extension of ADH on DVAB Removal
Lesion size
ADH extension
Benign
Malignant
Total
Complete
⬍6 mm
Complete
ⱖ6 mm
Incomplete
⬍21 mm
Incomplete
ⱖ21 mm
ⱕ2 ⬎2 ⱕ2 ⬎2 ⱕ2 ⬎2 ⱕ2 ⬎2
5 12 9 14 14 12 11 10 87
0 0 1 9 0 7 4 8 29
5 12 10 23 14 19 15 18 116
foci foci foci foci foci foci foci foci
ADH ADH ADH ADH ADH ADH ADH ADH
Total
DVAB ⫽ directional vacuum-assisted stereotactic biopsy; ADH ⫽ atypical ductal hyperplasia.
both the lesion size was below 6 mm and the lesion was completely removed. When the lesion size was 21 mm or greater, the incidence of malignancy increased dramatically (12/33). In these subgroups, the extent of ADH within the ducts had no impact (⬍6 mm and completely removed) or little impact (ⱖ21 mm) on underestimation: 4/15 in the ⱕ2 foci ADH group compared to 8/18 in the group with greater than 2 foci ADH. However, when the lesion size was ⱖ6 and less than 21 mm or when the lesion size was less than 6 mm and incompletely removed, the incidence of malignancy rose from 1/24 with ⱕ2 foci ADH (4%) to 16/42 with more than 2 foci ADH (38%).
Abstention group The follow-up was not available for 49/184 patients. The mean follow-up time for the remaining 135 patients was 30 months (range 6 months to 6 years). Four patients recurred (3%) within 1 to 5 years after initial diagnosis of ADH on DVAB. Table 6 reports the clinical and histological characteristics of these 4 tumors.
Table 6 Clinical and histological characteristics of recurrences in followed patients Lesion size
Initial biopsy
Time to recurrence
Unknown
ⱕ2 foci ADH
5 years
13 mm
ⱕ2 foci ADH
5 years
30 mm 4 mm
⬎2 foci ADH ⬎2 foci ADH
3 years 1 year
Surgical excision DCIS intermediate grade DCIS intermediate grade DCIS low grade Invasive ductal carcinoma
DCIS ⫽ ductal carcinoma in situ; ADH ⫽ atypical ductal hyperplasia.
Comments Various rates of ADH in DVAB are reported in the literature, ranging from 4% to 10%.2,4,6,8 –10,12,14 The overall ADH rate in our series was 14%, which is much higher than reported elsewhere. The difference might be due to difficulties in histological interpretation, patient selection bias, lesion size, and/or differences in needle-gauge as further described. Table 7 compares our findings with the largest published series of ADH on DVAB with regards to the number of procedures, the rate of ADH, the lesion size, the number of cores, the diameter of the needle (11/14 gauge), and the underestimation rates function of the final histological findings on surgical excision.2– 4,6,8 –10,12–14 Only 3 reports have specified the size of the lesion.8,10,12 In contrast to our series, both 11- and 14-gauge biopsies have been performed, which makes comparison difficult.8,10,12 Nevertheless, the mean size of the lesion in the current study seems to be almost twice as large as in other studies. This factor, together with the larger number of core biopsies (16 vs 11.5) and the larger diameter of the needle (11 gauge only) might explain the higher rate of ADH reported in our study. One could assume that the likelihood of ADH on DVAB might increase with more sampling, especially when larger lesions are removed. Only 2 of the other 9 studies have presented the follow-up of non-operated patients with evidence of ADH on DVAB. Both were conducted in France, where surgical excision after ADH on DVAB is not the rule. Indeed, 65% of the patients in these 2 series were spared surgery, which is similar to our population. For Plantade et al,4 the existence of one of the following criteria led to surgery: personal or family history of breast cancer, lesion size greater than 10 mm, presence of ADH on the last cores removed, or concomitant atypical lobular hyperplasia on DVAB. However, this study did not compare the outcomes of patients who underwent surgery and those who did not. In the Travade et al14 series, only patients with complete removal of the lesion by DVAB or those above 70 years of age who had no other risk factors did not undergo surgery. In our series, although personal or family history of breast cancer
(7) (7) (7) (4) (4) (6) (10) (6) (14)
ADH ⫽ atypical ductal hyperplasia; DVAB ⫽ directional vacuum-assisted stereotactic biopsy; IDC ⫽ infiltrative ductal carcinoma; DCIS ⫽ ductal carcinoma in situ; G1 ⫽ low grade; G2 ⫽ intermediate grade; G3 ⫽ high grade.
— — 0 (0) — 2 (2) — — — — — — 2 (2) — — 4 (9) — 8 (8) — — — — — — 5 (4) — — 11 (24) — 9 (8) — — — — — — 19 (16) (11) (4) (33) (14) (18) (5) (9) (13) (19) (19) (14) (22) 7 2 15 13 19 2 6 11 7 6 88 26 (3) (9) (3) (0) (3) (2) (8) (6) (8) (0) (4) (3) 2 4 2 0 3 1 5 5 3 0 25 3 (14) (13) (36) (14) (21) (7) (17) (19) (27) (19) (18) (25) 9/62 6/46 17/46 13/95 22/104 3/42 11/65 16/86 10/37 6/31 113/614 29/116 11 11 — 11/14 11 11/14 11 11/14 11 11 — 11 6 12 12 5 18 11 15 7 18 — 12 16 — — — 1–7 mm (2) — 3–85 mm (16) — 2–41 mm (3) — — 1–85 mm (9) 2–90 mm (18) (8) (5)
90 46 — 216 131 61 77 148 135 62 966 300 1,081 851 — 3,026 1,964 824 1,750 3,873 2,130 633 16,132 2,214
Mean no. of cores Lesion size (mean) No. of ADH (%) No. of DAVB
Needle core biopsy in atypical ductal hyperplasia
Adrales 20009 Burak 20002 Ely 200113 Renshaw 200110 Jackman 20023 Sneige 20038 Winchester 20036 Darling 200012 Plantade 20044 Travade 200614 Total Present study
Table 7
Underestimation of ADH at stereotactic DVAB: findings in the literature
Needle gauge
No. of underestimated/ surgery (%)
IDC (%)
DCIS (%)
DCIS G1 (%)
DCIS G2 (%)
DCIS G3 (%)
C. Forgeard et al.
343
and the type of radiological abnormality (microcalcifications associated or not with echographic mass) influenced the choice of abstention versus surgery, these data were not statistically different between the 2 groups of patients. Only the following criteria were statistically different between the 2 groups: lower average size of the lesion in the abstention group (12 vs 18, P ⬍.0001), more frequent complete removal of microcalcifications with DVAB (71% vs 43%, P ⬍.0001), and ⱕ2 foci ADH on biopsy (94% vs 38%, P ⬍.0001). Follow-up was performed by mammography at 3 months14 or at 6 months4,14 after DVAB, as was the case in our series. Plantade et al4 recommended further follow-up at 6 months and every year thereafter, whereas Travade et al14 proposed yearly mammography. The follow-up durations were 29 months (Plantade et al4) and 35.5 months (Travade et al14) in previous studies, and 30 months in ours. Interestingly, no event was noted in the series with the longest follow-up, whereas 2 invasive nonpleomorphic lobular carcinomas were observed in the study by Plantade et al4 (a condition not usually associated with microcalcifications). One can assume that ADH was a risk marker only in the vicinity of the tumor. Four events occurred in our series at, respectively, 5 years (1 intermediategrade DCIS), 3 years (1 low-grade and 1 intermediate-grade DCIS), and 1 year (one 8-mm Scarff-Bloom-Richardson grade 1 IDC) after the procedure (Table 6). All in situ carcinomas developed at least 3 years after ADH diagnosis, which may not indicate underestimation of malignancy by DVAB. The results might be related to ADH as a risk marker of breast cancer. The invasive carcinoma also may not be due to underestimation even though it happened within 1 year, because it occurred elsewhere in the breast and was suspected because of new microcalcifications. The initial lesion was only 4 mm and microcalcifications were completely removed by DVAB. Taken together, the data indicate that strict follow-up of ADH seems a safe alternative to surgery under certain circumstances (lesion ⬍10 mm, complete removal of microcalcifications by DVAB, and ⱕ2 foci ADH on DVAB). In contrast to “follow-up studies,” many authors have assessed the correlations between histological diagnoses on biopsies and surgical specimens. All have pointed out the underestimation risk. Initial studies using different needle diameters have shown underestimation rates between 11% and 67%. All have concluded that surgery must be performed in cases of ADH.1,3,5,6,11 More recent studies using DVAB and larger needles have reported lower underestimation rates of 0% to 38%.1,2,7,11,12 Table 7 shows the underestimation rates in the main series of the literature (10 recent series with more than 40 cases each).2– 4,6,8 –10,12–14 In our study, the underestimation rate for IDC was 3%, which is quite similar to average values (4%) from the main series in the literature. For DCIS, our underestimation rate was 22% compared to an average of 14% obtained from pooling the results of the main series. This discrepancy
344
The American Journal of Surgery, Vol 196, No 3, September 2008
might be due to low reproducibility in the diagnosis of intraductal proliferative lesions and/or to a patient selection bias. Nevertheless, the underestimation rate in the main series varied from 4% to 33% and our result is within this range. The grade of DCIS was specified only in 2 of the 8 series.3,13 The proportion of IDC is quite similar in the different series. In contrast, the distribution of DCIS grades seems to vary widely from one series to another, which is possibly due to the variable definition of DCIS in the literature. Either the diagnosis is based only on nuclear size or it also takes into account the presence or the absence of comedo necrosis, thus leading to poor diagnostic reproducibility.15–18 In all series, low-grade DCIS is the more frequent (8%–24%), followed by intermediate-grade DCIS (4%–9%) and high-grade DCIS (0%–2%). Since ADH and low-grade DCIS are morphologically very similar and distinguished only by the extension of the lesion, one can expect that low-grade DCIS represents the most frequent diagnosis (⬃60%) where underestimation occurs. Predictors of ADH underestimation include personal history of breast cancer,3,9 presence of palpable or ultrasonographic mass,11,19 large lesion size,3,10 incomplete removal of lesion with DVAB,3,4,8 –11,14,20 low number of cores,4,21,22 and ADH within at least 3 foci.8,9,11,13 In a large series of DVAB (N ⫽ 104), Jackman et al3 showed an important decrease in underestimation rate when patients had no personal history of cancer, a lesion size below 10 mm, and complete removal of the lesion. When these 3 criteria were met, the underestimation rate remained above 2% and the authors concluded that surgery was necessary. Renshaw et al10 in a large series (N ⫽ 95), but also Travade et al14 (N ⫽ 31) and Liberman et al20 (N ⫽ 10) in smaller studies, reported no underestimation when the lesion was totally removed by DVAB. In our series, the size of the lesion was higher in underestimated patients than in others (23 mm vs 16 mm). Moreover, when the lesion was ⱖ21 mm, the underestimation rate was high (36%), suggesting that it was probably due to unavoidable inadequate sampling of needle biopsies. In contrast, the relative risk of malignancy with a lesion size less than 21 mm was reduced (odds ratio .5, P ⬍.05), with an even lower risk in the group with lesions less than 6 mm (odds ratio .13, P ⬍.01). Indeed, we were able to demonstrate that below 6 mm, no malignancy was diagnosed on the surgical specimen when the lesion was completely removed by DVAB. Plantade et al4 found a significant decrease in the underestimation rate when the number of cores was at least 15. In our study, since the number of core biopsies was always high (average of 16), this factor could not come out. More recently several authors have described histological factors such as the extent of ADH on DVAB that would predict underestimation.8,9,13 Indeed, this particular factor was as a major criterion in our institution for choosing between surgery and follow-up: the proportion of patients with greater than 2 foci ADH was 28% in the total population compared to 62% in
the surgical group. In a series of 62 patients, Adrales et al9 also distinguished 2 populations of ADH with “mild or moderate atypia.” The risk of underestimation was approximately 9% in the first group and 44% in the second. However, these authors gave no definition of their diagnoses. Since underestimation was not observed in patients with both no personal or family history of breast cancer and mild atypia, they did not recommend surgical excision. Ely et al13 in a series of 46 patients distinguished 3 groups according to ADH extension: when ADH was confined to 2 foci or less no underestimation was seen. When ADH involved 3 foci or more than 4 foci, the underestimation rates were 50% and 87%, respectively. Since the underestimation rate increased proportionally with ADH extension, the authors concluded that when ADH is confined to 2 foci or less, surgical excision can be avoided. Sneige et al8 in a series of 42 patients found no underdiagnosis when ADH was confined to less than 3 foci and/or when the lesion was completely removed by DVAB. To our knowledge our series is the largest. The underestimation rate was 25% when the diagnosis of ADH without specification was made on DVAB. At our institution, since 1999 the extension of ADH within ducts on DVAB has been prospectively used as histological marker. In contrast to Ely et al13 our categories included only 2 levels: ⱕ2 foci ADH or greater than 2 foci ADH. Using these criteria, the underestimation rate in operated patients fell from 33% (⬎2 foci ADH) to 11% (ⱕ2 foci ADH). Moreover, in multivariate analysis the extent of ADH was the most important criterion when the size of the lesion was ⱖ6 mm and less than 21 mm and also with a lesion size less than 6 mm with incomplete removal with DVAB. In both cases, underestimation rates rose from 4% with ⱕ2 foci ADH to 38% with greater than 2 foci ADH. However, the extension of ADH on DVAB was not discriminating when the size of the lesion was less than 6 mm and completely removed (no underestimation) or when the size was ⱖ21 mm (36% total underestimation rate) with rates rising from 27% to 45% when ⱕ2 foci ADH or greater than 2 foci ADH were diagnosed, respectively.
Figure 1 Proposition of management for patients with ADH on directional vacuum-assisted stereotactic biopsy.
C. Forgeard et al.
Needle core biopsy in atypical ductal hyperplasia
Conclusion Although surgical excision is usually recommended for ADH on 11-gauge DVAB, recent data and results from this series tend to indicate that strict follow-up can be a safe alternative in a subset of patients representing almost 60% of the patients in the main series. In the present study, we have defined 3 criteria that can help in choosing between “surgery” or “follow-up.” These criteria include lesion size, complete removal of microcalcifications with DVAB, and extent of ADH. From these data, we deduced a schema of management for patients with ADH on DVAB as seen in Figure 1. In contrast to ⱖ21-mm lesions (right), surgery may be spared in cases with microcalcifications completely removed less than 6 mm (left). However, when the lesion size is less than 6 mm but with incomplete removal or when the size is ⱖ6 mm and less than 21 mm, abstention can be recommended for ⱕ2 foci ADH patients, whereas surgery seems mandatory in cases of greater than 2 foci ADH. We are currently using these new guidelines prospectively at our institution and will assess them in the future.
References 1. Zhao L, Freimanis R, Bergman S, et al. Biopsy needle technique and the accuracy of diagnosis of atypical ductal hyperplasia for mammographic abnormalities. Am Surg 2003;69:757– 62. 2. Burak WE Jr, Owens KE, Tighe MB, et al. Vacuum-assisted stereotactic breast biopsy: histologic underestimation of malignant lesions. Arch Surg 2000;135:700 –3. 3. Jackman RJ, Birdwell RL, Ikeda DM. Atypical ductal hyperplasia: can some lesions be defined as probably benign after stereotactic 11-gauge vacuum-assisted biopsy, eliminating the recommendation for surgical excision? Radiology 2002;224:548 –54. 4. Plantade R, Hammou JC, Fighiera M, et al. [Underestimation of breast carcinoma with 11-gauge stereotactically guided directional vacuumassisted biopsy]. J Radiol 2004;85:391– 401. 5. Cangiarella J, Waisman J, Symmans WF, et al. Mammotome core biopsy for mammary microcalcification: analysis of 160 biopsies from 142 women with surgical and radiologic followup. Cancer 2001;91:173–7. 6. Winchester DJ, Bernstein JR, Jeske JM, et al. Upstaging of atypical ductal hyperplasia after vacuum-assisted 11-gauge stereotactic core needle biopsy. Arch Surg 2003;138:619 –22. 7. Burbank F. Stereotactic breast biopsy of atypical ductal hyperplasia and ductal carcinoma in situ lesions: improved accuracy with directional, vacuum-assisted biopsy. Radiology 1997;202:843–7.
345
8. Sneige N, Lim SC, Whitman GJ, et al. Atypical ductal hyperplasia diagnosis by directional vacuum-assisted stereotactic biopsy of breast microcalcifications. Considerations for surgical excision. Am J Clin Pathol 2003;119:248 –53. 9. Adrales G, Turk P, Wallace T, et al. Is surgical excision necessary for atypical ductal hyperplasia of the breast diagnosed by Mammotome? Am J Surg 2000;180:313–5. 10. Renshaw AA, Cartagena N, Schenkman RH, et al. Atypical ductal hyperplasia in breast core needle biopsies. Correlation of size of the lesion, complete removal of the lesion, and the incidence of carcinoma in follow-up biopsies. Am J Clin Pathol 2001;116:92– 6. 11. Jacobs TW, Connolly JL, Schnitt SJ. Nonmalignant lesions in breast core needle biopsies: to excise or not to excise? Am J Surg Pathol 2002;26:1095–110. 12. Darling ML, Smith DN, Lester SC, et al. Atypical ductal hyperplasia and ductal carcinoma in situ as revealed by large-core needle breast biopsy: results of surgical excision. AJR Am J Roentgenol 2000;175: 1341– 6. 13. Ely KA, Carter BA, Jensen RA, et al. Core biopsy of the breast with atypical ductal hyperplasia: a probabilistic approach to reporting. Am J Surg Pathol 2001;25:1017–21. 14. Travade A, Isnard A, Bouchet F, et al. [Non-palpable breast lesions and core needle biopsy with Mammotome 11G: is surgery required in patients with atypical ductal hyperplasia?]. J Radiol 2006;87:307–10. 15. Tavassoli FA. Ductal carcinoma in situ: introduction of the concept of ductal intraepithelial neoplasia. Mod Pathol 1998;11:140 –54. 16. Page DL, Dupont WD, Rogers LW, et al. Atypical hyperplastic lesions of the female breast. A long-term follow-up study. Cancer 1985;55: 2698 –708. 17. Ghofrani M, Tapia B, Tavassoli FA. Discrepancies in the diagnosis of intraductal proliferative lesions of the breast and its management implications: results of a multinational survey. Virchows Arch 2006; 449:609 –16. 18. Lennington WJ, Jensen RA, Dalton LW, et al. Ductal carcinoma in situ of the breast. Heterogeneity of individual lesions. Cancer 1994;73: 118 –24. 19. Liberman L, Dershaw DD, Glassman JR, et al. Analysis of cancers not diagnosed at stereotactic core breast biopsy. Radiology 1997;203: 151–7. 20. Liberman L, Smolkin JH, Dershaw DD, et al. Calcification retrieval at stereotactic, 11-gauge, directional, vacuum-assisted breast biopsy. Radiology 1998;208:251– 60. 21. Jackman RJ, Burbank F, Parker SH, et al. Atypical ductal hyperplasia diagnosed at stereotactic breast biopsy: improved reliability with 14gauge, directional, vacuum-assisted biopsy. Radiology 1997;204: 485– 8. 22. Jackman RJ, Nowels KW, Shepard MJ, et al. Stereotaxic large-core needle biopsy of 450 nonpalpable breast lesions with surgical correlation in lesions with cancer or atypical hyperplasia. Radiology 1994; 193:91–5.