Intra-operative specimen analysis using faxitron microradiography for excision of mammographically suspicious, non-palpable breast lesions

Intra-operative specimen analysis using faxitron microradiography for excision of mammographically suspicious, non-palpable breast lesions

ARTICLE IN PRESS The Breast (2004) 13, 307–315 THE BREAST www.elsevier.com/locate/breast ORIGINAL ARTICLE Intra-operative specimen analysis using ...

255KB Sizes 0 Downloads 82 Views

ARTICLE IN PRESS The Breast (2004) 13, 307–315

THE

BREAST www.elsevier.com/locate/breast

ORIGINAL ARTICLE

Intra-operative specimen analysis using faxitron microradiography for excision of mammographically suspicious, non-palpable breast lesions M. Muttaliba,*, M. Tisdalla, R. Scawna, S. Shoushab, R.S. Cumminsa, H.D. Sinnetta a

Department of Breast Surgery, Charing Cross Hospital, Fulham Palace Road, London W6 8RF, UK Department of Histopathology, Charing Cross Hospital, London, UK

b

KEYWORDS Breast; Specimen; Microradiography

Summary Recent advances in digital imaging have made Faxitron microradiography an attractive alternative to intra-operative conventional specimen radiography (CSR) for the excision of wire-localized breast lesions. Faxitron specimen analysis time, usefulness of digital image manipulation and reexcision rates were evaluated in comparison to CSR in 299 consecutive wire-localized excisions for mammographically suspicious non-palpable breast lesions (172 procedures with Faxitron, 127 with CSR) in a non-randomized study. The corresponding mean operation times were 34.7 vs. 42.7 min and the respective reexcision rates were 19.8% vs. 31.5% (no significant difference on w2 analysis Po0:1). Faxitron digital image manipulation led to cavity biopsies in 50% (60/121) of the cancer excisions. In 19 of these (16%), histological excision margins were converted from incomplete to complete. The shorter Faxitron mean operating time enables an additional wire-localized operation per theatre list. Digital imaging guides the surgeon for additional cavity biopsies, resulting in re-excision rates as good as CSR. & 2004 Elsevier Ltd. All rights reserved.

Introduction The excision of mammographically suspicious nonpalpable breast lesions remains technically challenging. Lesions with microcalcifications can be localized using a guidewire under stereotactic imaging or directly biopsied with devices such as *Corresponding author. Tel.: þ 44-77-905-951-32; fax: þ 4420-876-11-888. E-mail address: [email protected] (M. Muttalib).

the mammotome or Advanced Breast Biopsy Instrumentation (ABBI). Ultrasound wire guidance can be used to localize echogenic targets. The application of industrial microradiography for clinical use, in the form of a Faxitron portable radiographic unit, has been described since the late 1960s.1 The parameters for specimen thickness, X-ray tube kilovoltage setting, and film type have been determined.2 The machine has been used on operative and autopsy specimens and also in processing gross material within hospital

0960-9776/$ - see front matter & 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.breast.2004.02.005

ARTICLE IN PRESS 308

M. Muttalib et al.

pathology departments. Key advantages included the unit’s portability in terms of its oven-size and mobility on wheels, and the ability to be connected to any standard electrical power outlet. Faxitron microradiography has also been used for research in the fields of Medicine,3,4 Dentistry,5 and Veterinary Science.6 In breast surgery, microradiographic and mammographic patterns of breast specimen clacifications have been characterized in relation to histopathological findings.7,8 Specimens were initially imaged directly onto radiographic film9,10 but the advent of digital image processing has enabled electronic image manipulation and storage. In the United Kingdom, there are currently 67 Faxitron units, with the majority (estimated 80%) used in pathology departments. Twenty-three units have digital imaging capability (data source from UK distributor: Quados, Berkshire, UK). We evaluated the potential advantages of intra-operative Faxitron digital imaging over conventional specimen radiography (CSR) in terms of time taken for specimen analysis, whether digital image manipulation altered intra-operative management (i.e. led to cavity biopsies being taken), and the reexcision rates as a consequence.

Patients and methods Women with mammographically suspicious impalpable breast lesions were retrospectively studied. Selection criteria included suspicious microcalcifications or parenchymal distortion; with core biopsies that were indeterminate (B3), suspicious for malignancy (B4) or malignant (B5) on histological scoring (Table 1, after Shousha11). The Faxitron MX-20 with DC-4 digital imaging workstation (Faxitron Corp., IL, USA) was acquired by a generous donation from Charing Cross League of Friends in November 2001 (Fig. 1). In the Faxitron group, excision procedures were per-

Table 1 Histological scoring of breast core biopsies (after Shousha11). Histological score

Interpretation

B1 B2 B3

Normal/inadequate Benign Benign, uncertain malignant potential Suspicious for malignancy Malignant

B4 B5

Figure 1 The Faxitron MX-20 with DC-4 digital-imaging workstation.

formed upon 171 consecutive patients with 172 hook-wires localized by stereotactic or ultrasound guidance over a 16-month period. The CSR group contained 126 consecutive patients with 127 wire guided excision procedures that were archived on the operating theatre database in the 32 months prior to Faxitron acquisition. At operation, the excised breast specimen was orientated by the surgeon with 2 silk marker sutures (short stitch superiorly, long stitch laterally) and placed into a clear plastic specimen bag. In the CSR group, the pre-operative mammograms and excised specimen were taken to the radiology department for standard compression radiography. The films were compared by a radiologist and then returned with the specimen and a comment regarding the completeness of excision. Further breast tissue was excised if required. In the Faxitron group, excised specimens were processed in the following manner. Inside the leadlined Faxitron X-ray cabinet, 2 orthogonal red laser markers denoted the centre of the X-ray detector window. The specimen was placed upon the

ARTICLE IN PRESS Intra-operative specimen analysis using faxitron microradiography

detector window at the bottom of the cabinet, and orientated to preserve the relationship of the marker sutures. The area, where the tip of the guidewire within the specimen, was expected to lie was placed at the intersection of the laser markers and the cabinet door closed. The cabinet door incorporates a safety switch which prevents radiation exposure if the door is ajar. A Windows ME (trademark, Microsoft Corp.) based workstation with proprietary Faxitron Imaging Software (Quados Ltd., Berkshire, UK) then initiates X-ray imaging at the click of the computer mouse. Calibration is automatic, and with the standard settings of 26 kV, 0.3 mA; the greyscale digital image is ready to view on the monitor as soon as the exposure time of 10 s has elapsed. These settings are user-adjustable if required. Images can be manipulated to enhance magnification, contrast and brightness. Multiple hard-copy radiographs can be printed using the Photographic Network Printer NP-1660 and V2 dry film (Codonics, OH, USA). Images can be stored onto hard disc drive or optical (CD-R/RW) media for retrieval. Digital image enhancement via the workstation is illustrated in the case of a 61-year-old patient with clustered microcalcifications in the right breast requiring stereotactic wire guided wide local excision. Histology from the excised specimen revealed low-grade ductal carcinoma in situ (DCIS). Figure 2 shows the standard unenhanced Faxitron image. A cluster of microcalcifications lie within the region bounded by the ‘‘V’’ of the hook-wire. The standard image can also be enhanced using the inverse video mode. Black and white are reversed like the negative from photographic camera film, where microcalcifications appear as black instead of white specks. The image intensity of the standard view can be adjusted, as demonstrated in Figs. 3 and 4. Image contrast can be progressively adjusted to exaggerate slight differences in specimen density. The microcalcifications are more prominent now and if the contrast level is continually adjusted to one

Figure 2 Standard unenhanced Faxitron image of microcalcifications.

309

Figure 3 Image intensity adjustment.

Figure 4 Contrast level adjustment to one extreme.

Figure 5 Magnification view of microcalcifications.

extreme, the area of interest is clearly seen within the wire ‘‘V’’. Selected areas of the standard image can be magnified. Fig. 5 shows a magnification view of the microcalcifications. The calcifications seen on the Faxitron specimen radiograph were compared to those seen in the preoperative mammogram. If the suspected lesion could not be seen or the microcalcification pattern continued to one edge of the specimen, then additional tissue (cavity biopsy) was excised to obtain clearance radiologically.

ARTICLE IN PRESS 310

Table 2

M. Muttalib et al.

Analysis of breast lesion type and method of wire guided localization.

Histological grading of lesion on core biopsy

Localizations in CSR group Ultrasound wire

B3 indeterminate B4 suspicious/B5 malignant Total

Stereotactic wire

Localizations in Faxitron group Total

Ultrasound wire

Stereotactic wire

Total

14 5

12 96

26 101

17 13

8 134

25 147

19

108

127

30

142

172

The histopathology reports for each patient’s surgically excised specimen were subsequently obtained and note was made of the histological type and grade of lesion, specimen weight, adequacy of margins, whether the patient required re-excision as a second operation, and whether intra-operative cavity biopsies were taken to facilitate complete excision. From the operating theatre records, the seniority of the surgeon and duration of operation were noted. The recorded data in the Faxitron and CSR groups were compared. The decision for a second (re-excision) operation was jointly made at a Multidisciplinary Team Meeting by consultants in Breast Surgery, Radiology, Histopathology, Oncology and Radiotherapy. Patients with inadequate margins (within 1 mm of cancer) would be offered a cavity re-excision. Patients with multifocal disease, situations where cavity re-excision would give a poor cosmetic result, or at the patient’s specific request; would be offered a completion mastectomy. Many malignant specimens had more than one cell type. To avoid specimens from being counted twice, the priority of invasive cancer4in situ cancer4benign was applied. Hence, a specimen with invasive ductal carcinoma (IDC) with surrounding ductal carcinoma in-situ (DCIS) would be classified in the IDC group only.

Results A total of 297 women with 299 procedures were included in the study. Mean age ¼ 57.5 years (range 33–80 years). The CSR group contained 126 patients with 127 procedures over a 32-month period. One patient had procedures on 2 separate occasions. The mean operating time was 42.7 min (range 27–81 min) in this group. The Faxitron group contained 171 patients with 172 procedures over a 16-month

period. One patient required bilateral breast wires simultaneously. The mean operating time in this group was 34.7 min (range 14–73 min). Table 2 shows that in the CSR group, there were 26 (20.5%) B3 indeterminate lesions. In the Faxitron group, there were 25 (14.5%) B3 lesions. In both groups, most ultrasound wire localizations for diagnostic biopsy were performed for B3 indeterminate lesions causing parenchymal distortion on mammography. In the CSR and Faxitron groups, there were 101 (79.5%) and 147 (85.5%) B4/B5 lesions, respectively; requiring therapeutic wirelocalized wide local excision. The majority of these were achieved under stereotactic wire guidance for suspicious mammographic microcalcifications. Table 3 shows that in the CSR group there were a total of 89 cancers, accounting for 70% of the 127 specimens. The benign:malignant ratio was 1:2.34. In the Faxitron group, there were a similar proportion of cancers; a total of 121 cancers accounting for 70% of the 172 specimens, with benign:malignant ratio of 1:2.37. The mean weight of excised specimen was comparable in the CSR and Faxitron groups, respectively, at 36.1 g (range 7–140 g) and 35.5 g (range 1.8–168.8 g). Table 4 shows that the re-excision rate when using the Faxitron is 19.8%, which is lower than CSR at 31.5%. Eighteen patients in the Faxitron group required cavity re-excision for incomplete histological margins despite microradiographic clearance. Ten of these patients were found to have residual tumour (i.e. failure of Faxitron to achieve complete excision). Six patients in the Faxitron group required completion mastectomy for multifocal disease. Four of these were found to have residual disease in the mastectomy specimen. By comparison in the CSR group, 10/21 patients with cavity re-excisions and 5/7 with completion mastectomies had residual tumour. w2 analysis of the data in Table 5 shows that there is statistically no significant difference (Po0:1) between the CSR and Faxitron groups in terms of re-excision rates. To see if the experience of the

Invasive cancer CSR Group

Invasive ductal carcinoma (IDC)

Grade 2

9

13

Ductal carcinoma in situ (DCIS)

Tubular Invasive carcinoma lobular carcinoma (ILC)

Grade 3

Low grade

7

29

24

7

Invasive ductal carcinoma (IDC)

Grade 1

Grade 2

19

43

Tubular Invasive carcinoma lobular carcinoma (ILC)

10 2 77

3 46

Grade 3

72

16

43

43 Faxitron group

Lobular Fibroadenoma, carcinoma in fibrocystic change, situ (LCIS) radial scar, atypical ductal hyperplasia

Intermediate High grade grade

3 7

Benign

3

Ductal carcinoma in situ (DCIS)

Low grade

Intermediate High grade grade

10

17

38 Lobular Fibroadenoma, carcinoma in fibrocystic change, situ (LCIS) radial scar, atypical ductal hyperplasia

12 39

5 44

51

ARTICLE IN PRESS

Grade 1

In situ cancer

Intra-operative specimen analysis using faxitron microradiography

Table 3 Analysis of excised specimen histopathology in the CSR and Faxitron groups.

311

ARTICLE IN PRESS 312

Table 4

M. Muttalib et al.

Breast cancer re-excision rates in the CSR and Faxitron groups. Type of re-excision Surgeon performing original wire localized excision

Cavity re-excision

Completion mastectomy

Total

CSR group

Consultant Registrar Total

19 2 21 (23.5%)

3 4 7 (7.8%)

22 (24.7%) 6 (6.7%) 28 (31.5%)

Faxitron Group

Consultant Registrar Total

15 3 18 (14.8%)

6 0 6 (4.9%)

21 (17.4%) 3 (2.5%) 24 (19.8%)

CSR group with 89 cancers, Faxitron group with 121 cancers.

Table 5

Statistical comparison of breast cancer re-excision rates between the CSR and Faxitron groups.

Re-excision required Excision complete Total

CSR

Faxitron

Total

28 (31.5%) 61 (68.5%) 89 (100%)

24 (19.8%) 97 (80.2%) 121 (100%)

52 (24.8%) 158 (75.2%) 210 (100%)

w2 test for association ¼ 3.72 with 1 degree of freedom (Po0.1).

Table 6

Usefulness of Faxitron digital image manipulation. Faxitron digital image appearances requiring cavity biopsy

Cases where further cavity biopsy was non-malignant on histology, confirming complete excision

Cases where further cavity biopsy helped to achieve clear margins, facilitating a complete excision

Cases where further cavity biopsy failed to achieve clear margins

28 (46.7%)

19 (31.7%)

13 (21.6%)

Total ¼ 60 (100%) Patients with multiple cavity biopsies were counted only once.

surgeon had any effect on the results, the data (Table 4) was adjusted to remove the operations performed by a registrar, so that consultantperformed operations were only considered. This again showed no significant difference between the two groups (Po0:1). All of the 172 specimens in the Faxitron group underwent digital image manipulation. Of the 121 cancer excisions in this group, half (60/121) had incomplete excision on digital imaging that required cavity biopsies (Table 6). Histological excision margins were converted from incomplete to complete in 19 of these cavity biopsies, enabling a 16% increase in complete cancer excisions.

Discussion The mean operating time in the Faxitron group was 34.7 min compared to 42.7 min in the CSR group. One hundred and seventy-two specimen analyses were performed in the 16-month Faxitron group study period (a rate of 10.75 per month) whereas in the preceding 33 months using CSR, there were 127 specimen analyses (a rate of 3.85 per month). This reflects the shorter mean operation time in the Faxitron group, as there is no requirement to transport specimens to the radiology department for analysis.

ARTICLE IN PRESS Intra-operative specimen analysis using faxitron microradiography

Since the Faxitron was introduced, operating list efficiency has improved by enabling 3 wire guided excisions per operating list whereas only 2 could previously be accommodated. This gain in efficiency justifies the initial financial outlay of d47,100 for the Faxitron MX-20 with digital workstation and printer. The annual cost of consumables (printer cassette tray containing 100 films) and biannual service support total d550. The Faxitron can print specimen radiographs for the patient’s records and also to assist the histopathologist in identifying calcifications on initial specimen sectioning.12 Indeed, one of the first uses of microradiography was to help process gross material within hospital pathology departments.1 In the Faxitron and CSR groups, the suspect breast lesion types and their method of wire guided localization are in similar proportions (Table 2). Likewise in both groups, the benign:malignant ratio of the excised specimens are similar (Table 3). This enables a valid comparison between the re-excision rates in the Faxitron and CSR groups. Our CSR re-excision rate of 31.5% (Table 4) is in line with previously reported research. In the Royal Marsden Hospital series of 151 needlelocalizations for cancer using CSR, 33% (50/151) required re-excisions for incomplete histological margins.13 Although the Faxitron re-excision rate of 19.8% is lower than the CSR re-excision rate of 31.5%, this difference is not statistically significant (Po0:1) on w2 analysis (Table 5). The Faxitron is therefore as good as CSR in terms of re-excision rates for incomplete histological margins. Faxitron digital image manipulation is useful because it guided the surgeon to take cavity biopsies that facilitated clearance in 19 out of 121 cancers (Table 6). This further increased the rate of complete excision by 16 to 80.2%. Cavity biopsies are known to increase the completeness of excision. In another study with 77 hook-wire guided wide local excisions for carcinoma, an 18% increase was reported.14 Ten out of 18 patients (55%) with cavity re-excisions in the Faxitron group were found to have residual disease. Cavity re-excision in another series showed residual cancer in 33% of 74 patients using a double-dye localization technique or ultrasound.15 Unsuccessful wire-localized excision has been associated with more than one lesion per breast, small lesions, small specimens, and microcalcifications.16 Failure was also associated with women aged under 55 years, and those with dense breasts (Wolfe grade DM or DY).17 The difficulty in achieving satisfactory localization is reflected in the variety

313

of alternative localization techniques that have been proposed.18–23 It is apparent from numerous studies24–29 that radiological clearance of excised specimens or their associated cavity biopsies are not sufficient in themselves to ensure complete tumour excision. This would also apply in the case of Faxitron microradiography, which guides the surgeon to excising more breast tissue, but is not a substitute for the final histopathology report. It is important to achieve clear histological margins for patients to avoid further surgery. Radiotherapy cannot compensate for inadequate surgery. In 418 patients over a 10-year period,30 where histologically, wide local excision was incomplete and patients received radiotherapy, the local recurrence rate was 17%. When using the Faxitron, it was often necessary to keep the excised specimen flat by using a plastic specimen bag and squeezing the air out. This can distort the tissue, but if this step is omitted, then X-rays may reflect from an air-plastic-air interface within an air pocket to cause a white-out image instead of X-rays being attenuated by an air– plastic–specimen interface. Techniques have been described for producing enhanced radiographs of excised breast specimens with conventional mammographic equipment. Specimen immersion in water improves visualization of tissue structure when compared to radiographs of the same specimen in air.31 Compression further improves water immersion images in up to 88% of cases.32,33 Further improvement has been obtained with 2 orthogonal specimen views.34 Specimen radiography in a tetrahedron has also been described35 but both techniques raise problems regarding maintaining the specimen in a fixed position. Differences between microcalcification morphology in surgically excised breast specimens have been further evaluated by microradiography combined with digital imaging techniques.36 Digitized microradiographs showed significant differences between fibrocystic change, benign and malignant tumours in terms of clustering parameters viz. number of particles per cluster, area of clusters, maximum distance to nearest neighbour, and geometric mean distance to nearest neighbour. The distribution pattern index (DPI) clustering parameter appears to be a useful radiological correlator with underlying pathology. A combination of water immersion and compression for breast specimens, with Faxitron image digitization and analysis of microcalcification clustering patterns, could improve radiological clearance of tumour margins, correlated with the

ARTICLE IN PRESS 314

histopathology report. We are performing further studies to evaluate this combination of techniques.

Conclusion The Faxitron is as good as CSR in terms of mean weight (volume) of tissue excised and subsequent reexcision rates for incomplete histological margins. It enables rapid per-operative radiological analysis of excised wire-localized breast lesions and their associated cavity biopsies. The shorter mean operating time enables an additional wire-localized operation per theatre list, justifying the initial financial outlay for the Faxitron unit. Digital image manipulation of the excised specimens led to cavity biopsies in half of the cases, and increased the number of complete excisions by 16% as a result.

Acknowledgements The authors would like to thank the following for their assistance in the preparation of this paper: P. Ireland for surgical theatre list data assistance. S. Muttalib for translation of cited paper in French. J. Lenten for translation of cited paper in German.

References 1. Fingerhut AG. A self-contained radiographic unit. Radiology 1968;90:1030. 2. Conlogue GJ, Marcinowski III F. Microradiography: a theoretical basis and practical applications. Radiol Technol 1987;58(4):301–9. 3. Reading AD, McCaskie AW, Gregg PJ. The inadequacy of standard radiographs in detecting flaws in the cement mantle. J Bone Joint Surg Br 1999;81(1):167–70. 4. Sondergaard G, Vetner M, Christensen PO. Prostatic calculi. Acta Pathol Microbiol Immunol Scand A 1987;95(3):141–5. 5. Moskow BS, Tannenbaum P, Bloom A. Visualization of the human periodontium using serial thin section contact radiography. J Periodontol 1985;56(4):223–33. 6. O’Callaghan MW, Pascoe JR, Tyler WS, Mason DK. Exerciseinduced pulmonary haemorrhage in the horse: results of a detailed clinical, post mortem and imaging study. IV. Changes in the bronchial circulation demonstrated by C.T. scanning and microradiography. Equine Vet J 1987;19(5): 405–10. 7. Hassler O. Microradiographic investigations of calcifications of the female breast. Cancer 1969;23(5):1103–9. 8. Rosen PP, Snyder RE, Robbins G. Specimen radiography for nonpalpable breast lesions found by mammography: procedures and results. Cancer 1974;34(6):2028–33. 9. Kurzl R, Baltzer J, Lohe KJ. Intraoperative specimen radiography in mamographically suspect, non palpable breast lesions. Experiences with the Faxitron unit. Fortschr Med 1979;97(38):1688–90 [in German].

M. Muttalib et al.

10. Kurzl R, Lohe KJ, Baltzer J, Druckmann R, Zander J. Radiography used during operations for biopsies of nonpalpable lesions of the breast which have been suspected because of mammography. J Gynecol Obstet Biol Reprod (Paris) 1981;10(5):431–4 [in French]. 11. Shousha S. Issues in the interpretation of breast core biopsies. Int J Surg Pathol 2003;11(3):167–76. 12. Rebner M, Helvie MA, Pennes DR, Oberman HA, Ikeda DM, Adler DD. Paraffin tissue block radiography: adjunct to breast specimen radiography. Radiology 1989;173(3):695–6. 13. Mokbel K, Ahmed M, Nash A, Sacks N. Re-excision operations in nonpalpable breast cancer. J Surg Oncol 1995;58(4): 225–8. 14. Dixon JM, Ravi Sekar O, Walsh J, Paterson D, Anderson TJ. Specimen-orientated radiography helps define excision margins of malignant lesions detected by breast screening. Br J Surg 1993;80(8):1001–2. 15. Tate JJ, Royle GT, McDonald P, Guyer PB, Taylor I. X-ray and ultrasound localization of non-palpable breast lesions and difficulties in management. J R Soc Med 1987;80(11): 678–80. 16. Jackman RJ, Marzoni Jr FA. Needle-localized breast biopsy: why do we fail? Radiology 1997;204(3):677–84. 17. Aitken RJ, MacDonald HL, Kirkpatrick AE, Anderson TJ, Chetty U, Forrest AP. Outcome of surgery for non-palpable mammographic abnormalities. Br J Surg 1990;77(6):673–6. 18. Fornage BD, Ross MI, Singletary SE, Paulus DD. Localization of impalpable breast masses: value of sonography in the operating room and scanning of excised specimens. Am J Roentgenol 1994;163(3):569–73. 19. Haj M, Kniaz D, Eitan A, Solomon V, Cohen I, Loberant N. Three years of experience with advanced breast biopsy instrumentation (ABBI). Breast 2002;8(5):275–80. 20. Smathers RL. Advanced breast biopsy instrumentation device: percentages of lesion and surrounding tissue removed. Am J Roentgenol 2000;175(3):801–3. 21. Gray RJ, Salud C, Nguyen K, Dauway E, Friedland J, Berman C, Peltz E, Whitehead G, Cox CE. Randomized prospective evaluation of a novel technique for biopsy or lumpectomy of nonpalpable breast lesions: radioactive seed versus wire localization. Ann Surg Oncol 2001;8(9):711–5. 22. Raptopoulos V, Baum JK, Hochman M, Karellas A, Houlihan MJ, D’Orsi CJ. High resolution CT mammography of surgical biopsy specimens. J Comput Assist Tomogr 1996;20(2): 179–84. 23. Cross MJ, Harms SE, Cheek JH, Peters GN, Jones RC. New horizons in the diagnosis and treatment of breast cancer using magnetic resonance imaging. Am J Surg 1993;166(6):749–53. 24. Lee CH, Carter D. Detecting residual tumor after excisional biopsy of impalpable breast carcinoma: efficacy of comparing preoperative mammograms with radiographs of the biopsy specimen. Am J Roentgenol 1995;164(1):81–6. 25. Graham RA, Homer MJ, Sigler CJ, Safaii H, Schmid CH, Marchant DJ, Smith TJ. The efficacy of specimen radiography in evaluating the surgical margins of impalpable breast carcinoma. Am J Roentgenol 1994;162(1):33–6. 26. Oakley KL, Going JJ. Specimen slice radiography of cancer in breast conserving excisions. J Clin Pathol 1995;48(11): 1028–30. 27. Chinyama CN, Davies JD, Rayter Z, Farndon JR. Factors affecting surgical margin clearance in screen-detected breast cancer and the effect of cavity biopsies on residual disease. Eur J Surg Oncol 1997;23(2):123–7. 28. Hasselgren PO, Hummel RP, Georgian-Smith D, Fieler M. Breast biopsy with needle localization: accuracy of

ARTICLE IN PRESS Intra-operative specimen analysis using faxitron microradiography

29.

30.

31.

32.

specimen X-ray and management of missed lesions. Surgery 1993;114(4):836–40. Stein MA, Karlan M. Immediate postoperative mammogram for failed surgical excision of breast lesions. Radiology 1991;178(1):159–62. Renton SC, Gazet JC, Ford HT, Corbishley C, Sutcliffe R. The importance of the resection margin in conservative surgery for breast cancer. Eur J Surg Oncol 1996;22(1):17–22. Eastgate RJ, Gilchrist KW, Matallana RH. Enhancement of tissue structure visualization in breast specimen radiography. Radiology 1979;132(3):744–6. Chilcote WA, Davis GA, Suchy P, Paushter DM. Breast specimen radiography: evaluation of a compression device. Radiology 1988;168(2):425–7.

315

33. Philip J, Harris WG, Rustage JH. Radiography of breast biopsy specimens. Br J Surg 1982;69(3):126–7. 34. Rebner M, Pennes DR, Baker DE, Adler DD, Boyd P. Two-view specimen radiography in surgical biopsy of nonpalpable breast masses. Am J Roentgenol 1987;149(2): 283–5. 35. Armstrong JS, Davies JD, Kulka J. Excision adequacy and characteristics of impalpable masses assessed by specimen mammography in a tetrahedron. Br J Surg 1994;81(3): 390–1. 36. Ng KH, Looi LM, Bradley DA. Microcalcification clustering parameters in breast disease: a morphometric analysis of radiographs of excision specimens. Br J Radiol 1996; 69(820):326–34.