BJR
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Verkooijen, H M
Right arrow Articles by van Vroonhoven, T J M V
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Verkooijen, H M
Right arrow Articles by van Vroonhoven, T J M V
British Journal of Radiology 74 (2001),1007-1012 © 2001 The British Institute of Radiology

Full paper

Risk factors for cancellation of stereotactic large core needle biopsy on a prone biopsy table

H M Verkooijen, MD, PhD1, P H M Peeters, MD, PhD2, I H M Borel Rinkes, MD, PhD1, R M Pijnappel, MD3, A Kaya1, W P Th M Mali, MD, PhD4 and T J M V van Vroonhoven, MD, PhD1

(on behalf of the COBRA study group)

1Department of Surgery, 2Julius Center for Patient Oriented Research and 4Department of Radiology, University Medical Center Utrecht, PO Box 85500, 3508 GA Utrecht, and 3Department of Radiology, Martini Hospital Groningen, PO Box 30033, 9700 RM Groningen, The Netherlands


   Abstract
Top
 Abstract
Introduction
 Patients and methods
 Results
 Discussion
 References
 
Among patients undergoing stereotactic needle biopsy of the breast on a prone biopsy table, a substantial proportion of planned procedures are terminated prematurely. This study was undertaken to identify risk factors for cancellation of these procedures and to derive a clinical rule for predicting cancellation. Risk factors for cancellation were assessed in a group of 476 consecutive patients with non-palpable lesions planned for large core needle biopsy. 64 (13%) of these planned procedures were cancelled. Multivariate regression analysis was applied to identify independent risk factors for cancellation. Validation took place by applying the logistic rule on a validation set, including 5 cancelled and 35 successful biopsy procedures. Mammograms that were difficult to interpret owing to extremely dense breast tissue, axillary location of the non-palpable lesion, body mass index below 20, less than 15 mm distance from the lesion to the chest wall or the presence of more than one non-palpable lesion were identified as independent risk factors. The logistic rule discriminated patients with successful and cancelled biopsy procedures with a receiver operator characteristic (ROC) area of 0.72. In the validation set, the area under the ROC curve was 0.92. The prediction rule, based on mammographic and clinical findings, discriminated patients with successful and cancelled needle biopsy procedures to a certain extent. The risk of cancellation of the stereotacic biopsy procedure is considerable in cases of very dense breast tissue or the presence of multiple risk factors.


   Introduction
Top
 Abstract
 Introduction
Patients and methods
 Results
 Discussion
 References
 
In the diagnosis of non-palpable breast disease, stereotactic large core needle biopsy is frequently applied as a safe, less expensive and patient friendly alternative to needle localized open breast biopsy [1]. During the course of a multi-institutional prospective study in The Netherlands evaluating diagnostic accuracy, patient satisfaction and cost effectiveness of stereotactic large core needle biopsy, it was observed that a considerable proportion (11%) of planned stereotactic biopsy procedures were terminated prematurely. This is in accordance with the findings of Phillpots et al [2], who reported that 16% of planned procedures were cancelled. A large proportion of cancelled procedures has an unfavourable effect on cost effectiveness, waiting lists and patient anxiety. We therefore set out to assess independent risk factors for cancellation of stereotactic guided large core needle biopsy procedures in an attempt to derive and validate a clinical prediction rule for cancellation.


   Patients and methods
Top
 Abstract
 Introduction
 Patients and methods
Results
 Discussion
 References
 
The present study was conducted in the context of the prospective Dutch multi-institutional COBRA (Core Biopsy after Radiological localization) study. The aim of the COBRA study was to evaluate diagnostic accuracy, patient satisfaction and cost effectiveness of stereotactic large core needle biopsy for non-palpable breast disease. Between 1997 and 2000, stereotactic large core needle biopsy and surgical excision of the non-palpable lesion was planned in 973 consecutive patients with 1029 non-palpable breast lesions. Patients with non-palpable breast lesions requiring histological examination were eligible for inclusion. Patients with a bleeding diathesis or taking anticoagulants and patients who were unable to maintain the prone position for 1 h were excluded from the study. The Dutch National Health Insurance Council and all local Institutional Review Boards approved the study protocol, and informed consent was obtained from all patients. The diagnostic work-up of all patients involved a standard study protocol. When large core needle biopsy showed in situ or invasive breast cancer, definitive surgical therapy was performed and included axillary dissection or sentinel node biopsy in cases with invasive cancer. When the large core needle biopsy did not show malignancy, needle localized open breast biopsy was carried out to ensure that no malignancy had been missed. The results of the study are reported elsewhere [3]. Biopsies were performed using a prone table (Fisher Imaging, Denver, CO; or Lorad Stereoguide, Danbury, CT), a 14 G core needle and a longthrow automated biopsy device with multiple passes (C R Bard Inc., Covington, GA). Digital stereotactic mammography was used to localize the lesions and specimen radiographs were obtained to confirm the presence of microcalcification. Radiologists performing the stereotactic biopsy procedures underwent special training; 10 biopsy procedures were attended and,subsequently, another 10 biopsy procedures were performed under the supervision of a radiologist with extensive experience in stereotactic breast biopsy.

Cancelled biopsy procedures were defined as procedures that were started but needed to be cancelled for any reason. All procedures classified as "cancelled" involved patients that actually entered the X-ray room and were positioned on the prone table. Planned procedures that were abandoned because the non-palpable lesion appeared too benign or non-existent were beyond the scope of this study.

We compared successful and cancelled biopsy procedures for a number of variables. The following data were prospectively obtained: age, history of breast cancer, history of breast surgery, history of radiotherapy and body mass index (BMI) (weight (kg)/height2 (m2)). Family history of breast cancer was registered and defined as the occurrence of breast cancer in grandmother, mother, sister or daughter. The number of non-palpable lesions planned for large core needle biopsy, caudal or axillary location of the lesion, presence of microcalcification (calcified vs non-calcified lesion) and distance of the lesion from the chest wall in millimeters (measured from the centre of the lesion to the rim of the mammogram, estimated on craniocaudal mammograms) were also recorded. Breast density pattern was categorized as: (1) mostly fatty, (2) scattered fibroglandular densities, (3) heterogeneously dense and (4) extremely dense breast tissue. Breast size was defined as small breast size (cup size <=AA/A), moderate breast size (cup B/C) or large breast size (cup size >=D/DD).

Successful and cancelled procedures were compared for the aforemetioned variables in two sets of patients. The derivation set consisted of 476 consecutive patients planned for stereotactic large core needle biopsy between June 1997 and June 1999 and the validation set consisted of 40 consecutive patients planned for large core needle biopsy between July and September 1999.

Data analysis
In deriving and validating the prediction rule, continuous variables were tested with the Student's t-test for normally distributed data or the Mann-Whitney U-test for data not distributed normally. If a significant difference was found in a continuous variable, the variable was divided into categories. Odds ratios with 95% confidence intervals (CIs) were used to estimate relative risks of categorical variables for cancellation of biopsy procedure. Within the derivation set, all univariate correlates with a p-value <0.15 were entered into a stepwise logistic regression procedure. Diagnostic information or discriminative value of the logistic model was evaluated using the area under the receiver operating characteristic (ROC) curve [4]. Model reduction was performed by comparing areas under the ROC curves to obtain the most efficient diagnostic model. As all models were based on the same subjects, their correlation was taken into account [5].

Adjusted odds ratios with their 95% CIs were derived from the estimated regression coefficients. In this way, independent predictors for cancellation of the biopsy procedure were estimated and a prediction rule was derived.

The reliability of the model was evaluated by applying the logistic rule to the validation set. Discrimination of patients with and without cancelled procedures was again measured by ROC analysis. Sensitivities and specificities were calculated according to standard methods. Exact CIs of 95% were calculated.


   Results
Top
 Abstract
 Introduction
 Patients and methods
 Results
Discussion
 References
 
64 (13%) of 476 scheduled procedures in the derivation set were terminated prematurely. 58% of the lesions for which the biopsy procedure was cancelled proved to be malignant (ductal carcinoma in situ or invasive breast cancer), which is comparable with the malignancy ratio of successful procedures. Table 1Go presents the number of planned procedures and the proportion of cancelled procedures per hospital. Location of the lesion close to the chest wall and negative stroke margin, i.e. inadequate space for needle throw, were reported to be the most important reasons for cancellation (Table 2Go). The proportion of cancelled biopsy procedures remained stable over time (data not shown).


View this table:
[in this window]
[in a new window]
  		Table 1. Number of planned biopsy procedures and proportion of cancelled procedures in three participating hospitals

 

View this table:
[in this window]
[in a new window]
  		Table 2. Reasons for cancellation of stereotactic large core needle biopsy procedure reported by radiologists

 
Univariate analysis of continuous variables (age, distance of the lesion from the chest wall and BMI) demonstrated no difference in age between patients with successful or cancelled biopsy procedures. However, patients with cancelled procedures were leaner, i.e. had a significantly lower BMI, and had a smaller distance from the lesion to the chest wall. These variables were classified into categories. BMI was classified in two categories (<20 and >=20) and lesion distance from the chest wall was classified in two categories (<=15 mm and >15 mm). Of the categorical risk factors, axillary or caudal location of the lesion, BMI <20, distance of lesion from the chest wall <=15 mm, breast density pattern, presence of more than one lesion and small breast size were significantly associated with an increased risk of cancellation of the biopsy procedure (Table 3Go).


View this table:
[in this window]
[in a new window]
  		Table 3. Univariate analysis of risk factors

 
The multivariate analysis was based on 431 patients because 46 patients were excluded owing to missing values. The most efficient model on multivariate analysis included axillary location of the lesion, BMI <20, distance of the lesion from the chest wall <=15 mm, mammograms difficult to interpret owing to extremely dense breast tissue and the presence of more than one lesion. Adjusted odds ratios of these significant independent predictors for cancellation of stereotactic biopsy procedures are presented in Table 4Go. The area under the ROC curve for the logistic rule was 0.72 (95% CI 0.64–0.80). This value can be interpreted as a probability of 72% that the logistic rule would correctly identify patients with and without cancelled biopsy procedures. When applied to the validation set, the logistic rule discriminated patients with and without cancelled procedures with a ROC area of 0.92 (95% CI 0.83–1.02).


View this table:
[in this window]
[in a new window]
  		Table 4. Multivariate predictors of cancellation of biopsy procedures with logistic coefficients, odds rations, 95% confidence interval (CI) and score (rounded off coefficients)

 
For clinical purposes, a score system was made by simplifying the logistic model after rounding off the logistic coefficients (Table 4Go). Accordingly, a score based on the presence of one or more risk factors could be calculated for every patient. One point was added for the presence of more than one lesion and if the distance from the lesion to the chest wall was <=15 mm; 2 points for a BMI below 20 and if the lesion was located in the axilla; and 3 points for difficult to interpret mammograms owing to extremely dense breast tissue. This modified logistic rule discriminated patients with and without cancelled procedures with a ROC area of 0.72 (95% CI 0.64–0.80) in the derivation set and 0.93 (95% CI 0.84–1.02) in the validation set.

The prevalence of cancelled biopsy procedures according to cut-off score, and the sensitivity and specificity of the logistic rule for different cut-off points, are presented in Table 5Go. 50% of the scheduled biopsy procedures were cancelled (44% in the validation set) in patients with a score of 2 or more. 38% of the cancelled procedures in the derivation set would have been prevented if patients with a score of 2 or more had not been referred for needle biopsy. However, 5.9% of lesions in the derivation set would not have been referred for stereotactic large core needle biopsy, although successful needle biopsy would have been possible. 73% of the biopsy procedures were cancelled in patients with a score of 3 or more. 28% of the cancelled procedures would have been prevented if this cut-off score of 3 had been applied, while 1.6% of the potential successful biopsy procedures would not have been undertaken.


View this table:
[in this window]
[in a new window]
  		Table 5. Prevalence of cancelled biopsy procedures, potential number of cancelled procedures prevented and number of procedures incorrectly classified as unsuccessful, varying with different cut-off scores in the derivation and validation sets

 

   Discussion
Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
References
 
Risk factors for cancellation of stereotactic large core needle biopsy procedures on a prone biopsy table were identified in the present study. Location of the lesion close to the chest wall or in the axilla, the presence of more than one non-palpable lesion scheduled for stereotactic needle biopsy, mammograms that were difficult to interpret owing to extremely dense breast tissue and low BMI significantly and independently increased the probability of cancellation of the biopsy procedure. A clinical prediction rule based on these five findings classified cancelled biopsy procedures and successful biopsy procedures in the derivation set with a ROC curve area of 0.72. When applied to a patient sample other than that from which it was derived, the prediction rule maintained its discriminatory ability (area under ROC curve 0.92). An area under the ROC curve of more than 0.70 is generally considered as good, and an area of more than 0.80 as excellent.

Location of the lesion in the axilla or close to the chest wall were described earlier as complicating factors for stereotactic needle biopsy on a prone biopsy table [2, 6]. However, 63 (73%) of the 86 patients with lesions <=15 mm from the chest wall were actually able to undergo the needle biopsy procedure. We therefore consider that location of the lesion close to the chest wall is, in itself, not a reason for excluding patients from stereotactic biopsy on a prone table, unless other risk factors for cancellation are present.

Non-palpable lesions may be difficult to identify on mammography in cases of extremely dense breast tissue. This may lead to difficulty in identifying the lesion on stereotaxis. Ultrasound guidance is probably more appropriate for the visualization of most of these lesions. Ultrasound guided large core needle biopsy should be used as the method of choice for image guided biopsy of ultrasound visible lesions, because of lower costs and patient comfort. However, the aim of the COBRA study, in the context of which the current study was executed, was to assess the diagnostic accuracy of stereotactic large core needle biopsy. The study protocol therefore required the application of stereotactic guidance for every non-palpable lesion that was visible on mammography.

The presence of more than one non-palpable breast lesion lengthens the time of the biopsy procedure and thereby increases the risk of patients being unable to tolerate the relatively uncomfortable prone position. Nevertheless, large core needle biopsy is extremely useful for the diagnosis of multiple non-palpable breast abnormalities. For example, mastectomy can be planned if multiple carcinomas are detected in one breast. We therefore advocate not to exclude patients with more than one non-palpable lesion for large core needle biopsy, unless other risk factors are present or unless the patient is reluctant to undergo stereotactic breast biopsy.

On univariate analysis, low BMI and small breast size were both significantly associated with cancellation of the stereotactic biopsy procedure. Cancellation in the category of patients with small breasts might occur owing to either negative stroke margin or because the breast is too small for adequate compression and visualization of the lesion. However, on multivariate analysis, small breast size proved not to be an independent risk factor. An explanation for this might be that low BMI and small breast size are often associated. Extremely obese women are also at higher risk of cancellation of the biopsy procedure, because prone biopsy tables do not allow patients weighing over 135 kg. No women weighing over 135 kg were encountered in the present study.

All radiologists in this study underwent special training before performing stereotactic biopsy procedures. However, there was some variation in the level of experience in the use of the stereotactic biopsy technique among radiologists. But, as the proportion of cancelled biopsy procedures remained stable over time, the level of experience of the radiologist seems not to be associated with the risk of cancellation.

The results of this study suggest that it is possible to some extent to identify patients in whom the needle biopsy procedure will not be successful, based on the findings at mammography and physical examination. However, because the prediction rule does not discriminate perfectly, one might argue over the usefulness of such a rule and the ideal cut-off score. A large number of cancelled biopsy procedures would be prevented if, for example, all patients with at least one risk factor (cut-off score of 1) were considered inappropriate candidates for stereotactic needle biopsy. On the other hand, a large proportion of potentially successful biopsy procedures would not have been carried out either. More potentially successful biopsy procedures would occur if a higher cut-off score, for example the presence of at least two risk factors or the presence of one risk factor with a high score, were taken. However, prevention of procedures being cancelled would also decline.

Introduction of stereotactic large core needle biopsy is expected to reduce the costs associated with diagnosis of non-palpable breast disease, by replacing a substantial proportion of needle localized open breast biopsies [7–9]. However, a large number of cancelled biopsy procedures will have a negative impact on this cost reduction. A cost assessment in the context of the COBRA study demonstrated that a successful large core needle biopsy procedure costs {euro}550 [9]. The cost of a cancelled biopsy procedure was {euro}119 and the cost of open breast biopsy was {euro}1184. {euro}119 are therefore saved when cancellation of one stereotactic large core needle biopsy procedure is prevented. On the other hand, open breast biopsy is performed if a lesion that could have successfully been approached by needle biopsy is incorrectly not planned for stereotactic needle biopsy. Consequently, the extra cost of incorrectly not scheduling these patients for needle biopsy is {euro}634 (cost of open biopsy {euro}1184)-cost of needle biopsy ({euro}550)). Thus, approximately 5.4 cancelled biopsy procedures need to be prevented for every patient incorrectly not scheduled for stereotactic large core needle biopsy, which would be the case with a cut-off score >=4. With this cut-off point, only two cancelled procedures would have been prevented in our study population.

Nevertheless, cost reduction is not the sole objective for reducing the number of cancelled biopsy procedures. A large proportion of cancelled procedures might also unfavourably influence patient anxiety. Cancellation of the biopsy procedure induces diagnostic delay, which has a negative impact on psychiatric morbidity among women attending breast cancer screening [10].

Biopsy procedures that were not initiated, because the non-palpable lesions were re-assessed as benign before starting the needle biopsy procedure, were beyond the scope of this study. However, during the course of the present study, 14 biopsy procedures were cancelled for this reason. This was also the main reason for cancellation in the study of Philpotts et al [2]. It is therefore of utmost importance that complete mammographic work-up (including ultrasound examination, magnification and spot compression views) is performed before a patient is scheduled for stereotactic large core needle biopsy.

In conclusion, a prediction rule based on mammographic and clinical findings in patients with non-palpable breast lesions discriminated patients with successful and cancelled needle biopsy procedures to a certain extent. The risk of cancellation of stereotacic biopsy procedures is particularly high in cases of very dense breast tissue or if multiple risk factors are present. In addition, factors such as difficult location of the lesion or multiple lesions that are likely to result in a long procedure should be taken into account by staff when planning appointment schedules.


   Footnotes
 
Current address for H M Verkooijen: Geneva Cancer Registry, Institute for Social and Preventive Medicine, Boulevard de la Cluse 55, 1205 Geneva, Switzerland. Back

The present study was funded by the Dutch Ministry of Public Health. Back

The COBRA study group includes APE Besnard, IHM Borel Rinkes, E Buskens, WAH Gelderman, AD Groote, JHCL Hendriks, R Holland, PKH Hut, HJ de Koning, WPThM Mali, M Oudkerk, PHM Peeters, JL Peterse, RM Pijnappel, EJT Rutgers, MEI Schipper, P Schipper, JG van den Tweel, HM Verkooijen, ThJMV van Vroonhoven, PFGM van Waes and T Wiggers. Back

Received for publication March 6, 2001. Revision received June 11, 2001. Accepted for publication June 26, 2001.


   References
Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 

  1. Jackman RJ, Nowels KW, Rodriguez-Soto J, Marzoni FA, Finkelstein SD, Shephard MJ. Stereotactic, automated, large-core needle biopsy of nonpalpable breast lesions: false-negative and histologic underestimate rates after long-term follow up. Radiology 1999;210:799–805.[Abstract/Free Full Text]
  2. Philpotts LE, Lee CH, Horvath LJ, Tocino I. Canceled stereotactic core-needle biopsy of the breast: analysis of 89 cases. Radiology 1997;205:423–8.[Abstract/Free Full Text]
  3. Verkooijen HM. Stereotactic large-core needle biopsy for nonpalpable breast disease: the COBRA-study. Utrecht, The Netherlands: Utrecht University, 2000. [Thesis.]
  4. Hanley JA, McNeil BJ. The meaning and use of the area under a receiving operating characteristic (ROC) curve. Radiology 1982;143:29–36.[Abstract/Free Full Text]
  5. Hanley JA, McNeill BJ. A method of comparing the areas under receiver operating characteristic curves derived form the same cases. Radiology 1983;148:839–43.[Abstract/Free Full Text]
  6. Scott Soo M, Walsh R, Patton J. Prone table stereotactic breast biopsy: facilitating biopsy of posterior lesions using the arm-through-the-hole technique. AJR 1998;171:615–7.[Free Full Text]
  7. Howisey RL, Acheson MB, Rowbotham RK, Morgan A. A comparison of medicare reimbursement and results for various imaging-guided breast biopsy techniques. Am J Surg 1997;173:395–8.[Medline]
  8. Hrung JM, Langlotz CP, Orel SG, Fox KR, Schnall MD, Schwartz JS. Cost-effectiveness of MR imaging and core-needle biopsy in the preoperative work-up of suspicious breast lesions. Radiology 1999;213:39–49.[Abstract/Free Full Text]
  9. Buijs-van der Woude T, Verkooijen HM, Pijnappel RM, Klinkenbijl JHG, Borel Rinkes IHM, Peeters PHM, et al. Cost comparison between stereotactic core needle biopsy versus surgical excision biopsy in the Netherlands. Eur J Cancer 2001;37:1736–45.
  10. Ellman R, Angeli A, Christians A, Moss S, Chamberlain J, Maguire P. Psychiatric morbidity associated with screening for breast cancer. Br J Cancer 1989;60:781–4.[Medline]



This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
R. J. Jackman and F. A. Marzoni Jr.
Stereotactic Histologic Biopsy with Patients Prone: Technical Feasibility in 98% of Mammographically Detected Lesions
Am. J. Roentgenol., March 1, 2003; 180(3): 785 - 794.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Verkooijen, H M
Right arrow Articles by van Vroonhoven, T J M V
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Verkooijen, H M
Right arrow Articles by van Vroonhoven, T J M V

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
BJR DMFR IMAGING  ALL BIR JOURNALS