US correlation for MRI-detected breast lesions in women with familial risk of breast cancer.

AIM: To examine the value of US correlation for MRI-detected breast lesions in women with familial risk of breast cancer. METHODS: From an initial dataset of 245 women with positive family history who had breast cancer surveillance involving mammography or MRI between November 1994 and February 2001, 179 subjects with follow-up data were selected. A total of 43 women with 48 MRI-detected lesions underwent further assessment with US. Histopathological correlation was available from 38 breast biopsies performed for 33 women. RESULTS: Sonographic correlates were identified in 32 (66.7%) of the 48 MRI-detected lesions, with cancer present in 11 (34.4%) of these. This compares with 1 (6.3%) cancer found in the 16 lesions without sonographic correlates. Of the 12 malignant lesions, 11 (91.7%) had sonographic correlates whereas 21 (58.3%) of the 36 benign lesions had sonographic correlates. In all 74% of breast biopsies were performed under US guidance compared with 8% under MRI guidance. The proportion of MRI- and US-correlated benign and malignant lesions undergoing US-guided biopsy were 85.7% and 90.9%, respectively. CONCLUSION: The probability of cancer was significantly higher in MRI-detected breast lesions with sonographic correlates compared with those without such correlation. The advantage of convenient biopsy under US guidance as opposed to MRI guidance highlights the value of sonographic assessment of MRI-detected breast lesions.

Breast ultrasound in women with familial risk of breast cancer.

INTRODUCTION: The aim of this study was to assess the performance and value of breast ultrasound in women with familial risk of breast cancer. MATERIALS AND METHODS: From an initial dataset of 245 women with positive family history who had breast cancer surveillance utilising mammography or magnetic resonance imaging (MRI) between November 1994 and February 2001, 179 subjects with follow-up data were selected. Eighty-four women had breast ultrasound done with histopathological correlation available from 48 breast biopsies performed in 42 women. RESULTS: The sensitivity of ultrasound, mammography and MRI was 83.3%, 53.9% and 93.3%, respectively. The specificity of ultrasound, mammography and MRI was 65.5%, 85.7% and 63.6%, respectively. Ultrasound was the imaging modality with intermediate sensitivity, specificity, negative predictive value (NPV) and cancer detection rate. The sensitivity, specificity, positive predictive value (PPV), NPV and accuracy of combined mammography and ultrasound were 92.9%, 62.5%, 52.0%, 95.2% and 71.7%, respectively. These results did not differ significantly from MRI. Almost two-thirds of the breast biopsies were performed under ultrasound guidance. CONCLUSIONS: Although breast ultrasound screening per se was not assessed in this study, extrapolation of these results to sonographic screening of high familial risk women would come at a better specificity to MRI, albeit with a 10% decrease in sensitivity but at a fraction of the cost of MRI. Ultrasound also provides the advantage of convenient imaging guidance for biopsy. Employing ultrasound following mammography would match MRI in sensitivity, specificity, PPV, NPV and accuracy, and should not be ignored in these women.

The significance of circumscribed malignant mammographic masses in the surveillance of BRCA 1/2 gene mutation carriers.

Breast cancers in gene mutation carriers may escape mammographic detection because of rapid growth and tumor expansion. Therefore, they may mimic benign lesions on the mammogram. Twenty-nine BRCA 1/2 mutation carriers under surveillance developed 31 breast cancers between 1994 and 2001 at a mean age of 44.2 years. Controls were 63 women with 67 breast cancers in the same period at a mean age of 53.8 years, also under surveillance because of a life time risk of at least 15%. In 26% of the carriers vs. 48% of the controls, mammography was the method that first suspected a malignancy. Seven radiologists performed a retrospective review of the original mammograms to establish technical assessment, with special attention for circumscribed lesions and estimated probability of malignancy. In the mutation carriers seven (23%) circumscribed non-calcified mammographic masses were found and three in the controls (4.5%) P=0.01. These masses were proven to be malignant. In both groups around 70% of these fast-growing circumscribed lesions were detected by the patients. The masses were situated in breasts with a good interpretable breast pattern. BRCA 1/2 mutation carriers had a significantly higher percentage of circumscribed non-calcified mammographic masses that proved to be malignant. These mammographic lesions in women at high risk should be described as at least Birads 0 and worked-up with ultrasound and needle biopsy.

Reproducibility of mammographic classifications for non-palpable suspect lesions with microcalcifications.

Observer variability in interpretation of mammograms is a well-known problem, especially for microcalcifications. The classification of the mammographic findings depends upon this interpretation. We performed an intraobserver study to evaluate a breast imaging reporting and data system (BI-RADS) based method for description and classification of non-palpable lesions with microcalcifications. A set of 100 non-palpable mammographic lesions mainly consisting of microcalcifications was described and classified on two occasions, by two radiologists at an interval of 6 months. The intraobserver variability was evaluated with kappa statistics. The overall agreement for the classification was moderate (kappa 0.54). The lowest kappa values were observed for the categories "probably benign (BI-RADS 3, kappa 0.59)" and "suspicious abnormality (BI-RADS 4, kappa 0.44)". The clinical management (follow-up or biopsy) of non-palpable lesions consisting of microcalcifications depending upon radiological classification in the groups BI-RADS 3 (follow-up) and BI-RADS 4 (biopsy) is therefore debatable.

Diagnostic accuracy for different strategies of image-guided breast intervention in cases of nonpalpable breast lesions.

To find out whether ultrasound-guided fine-needle aspiration (FNA) and ultrasound and stereotactic-guided large core needle biopsy (LCNB) are reliable alternatives to needle-localised open breast biopsy (NLBB) in daily practice, we performed a retrospective study and evaluated the validity of these methods. In all, 718 women with 749 nonpalpable breast lesions from three Dutch Hospitals were included, and the validity of the various methods for diagnosis was assessed. This was carried out according to a method described by Burbank and Parker for evaluating the quality of an image-guided breast intervention. We compared our results with the outcome of the COBRA study. Overall, all diagnostic strategies (NLBB, FNA, LCNB ultrasound and stereotactic guided) show comparable agreement rates. However, the miss rates differ: 2% for NLBB, 3% for COBRA (LCNB in study setting), 5% for FNA and 8-12% for LCNB in practice. Fine-needle aspiration was nonconclusive in 29%, and shows an overestimation for DCIS in 9%. The DCIS underestimate rate in NLBB was 8%. For the assessment of lesions consisting of microcalcifications only and to exclude malignancy in all other lesions, a 14-gauge needle should be used. Ultrasound-guided intervention can be performed in a large percentage of nonpalpable lesions. Lesions consisting only of microcalcifications on mammography need special attention.

[The treatment of ductal carcinoma in situ (DCIS) of the breast]. / De behandeling van ductaal carcinoma in situ (DCIS) van de mamma.

Ductal carcinoma in situ (DCIS) of the breast is diagnosed more and more often in the Netherlands as a result of mammographic population screening and improved mammography techniques. Mastectomy and local excision, with or without radiotherapy, are used for the treatment of DCIS, but breast-conserving therapy seems a logical option in view of the favourable prognosis. Radiotherapy following total excision of DCIS reduces the local recurrence rate by half. Incomplete excision of DCIS is associated with a higher rate of local recurrence, which is invasive in about 50% of the cases and can therefore affect the prognosis adversely. There are conflicting opinions in the literature as to which patients with DCIS can be treated with breast-conserving therapy and whether local excision should be followed by radiotherapy, as is the case for breast-conserving treatment of invasive mammary carcinoma. The diagnosis and treatment of DCIS are therefore complex and require a multidisciplinary approach. Patients may be selected for breast-conserving therapy on the basis of diagnostic characteristics and risk factors.

Interobserver variability between general and expert pathologists during the histopathological assessment of large-core needle and open biopsies of non-palpable breast lesions.

The purpose of this study was to assess whether general pathologists are able to make as accurate and reproducible a diagnosis on large-core needle biopsies as on open breast biopsy specimens. A total of 688 patients underwent a stereotactic large-core (14G) needle biopsy and subsequent surgical excision of 718 non-palpable breast lesions. Forty-two pathologists from 10 departments of pathology (generalists) made a diagnosis on both the needle and open biopsy specimens. Afterwards, three pathologists and two radiologists with extensive experience in breast pathology (experts) diagnosed all of the biopsy specimens. The general pathologists made a similar histological diagnosis as the experts in 632 (88%) of the needle biopsies and 649 (90%) of the open biopsy specimens. Accordingly, the interobserver agreement for the diagnosis of large-core needle biopsies between the general and experts pathologists was excellent (kappa 0.83) and not significantly different from the interobserver agreement for the diagnosis of open breast biopsies (kappa 0.86). However, many inconsistencies were observed in the category of borderline lesions: only 24% of the large-core needle biopsies and 43% of the open biopsies with an expert diagnosis of 'borderline' were diagnosed similarly by the general pathologists. Additionally, the risk of benign/malignant inconsistencies between general pathologists and experts was approximately 1 in 55 for both needle and open biopsies.

Use of previous screening mammograms to identify features indicating cases that would have a possible gain in prognosis following earlier detection.

False-negative screening mammograms generally refer to breast cancers that were overlooked or misinterpreted at screening. An important question is whether earlier detection could have made a difference in the prognosis of the women concerned. We reviewed screening and diagnostic mammograms of 234 screen-detected and interval cancer cases (aged 44-84 years) diagnosed between 1991 and 1996 in the Nijmegen breast cancer screening programme. A lesion was visible on 117 (50%) of the screening mammograms prior to the diagnosis of breast cancer. Fifty-one out of the 117 cancers had poor prognostic characteristics at diagnosis (i.e. N+ and/or T2+) and could potentially have benefited from an earlier diagnosis ('possible gain'). The 'possible gain' cases were more often characterised by architectural distortion (29 vs. 10%; P=0.01) or a high-density mass (25 vs. 13%; P=0.06) on the mammogram prior to diagnosis than the 58 'no gain' cases. Our study shows that architectural distortion and non-spiculated high-density masses on the mammogram prior to diagnosis are associated with a possible gain in prognosis. Earlier detection of the carcinomas preceded by these signs may well have an impact on breast cancer mortality and thus warrant extra attention in radiological practice.

Diagnostic strategies in non-palpable breast lesions.

The number of non-palpable breast lesions is growing. Needle-localised breast biopsy (NLBB) is the gold standard for evaluating these lesions. Cost-saving techniques and less invasive alternatives such as core-needle biopsy (LCNB) and fine-needle aspiration (FNA) have emerged. The aim of this study was to find out if the lesions of patients who were sent directly for surgery to undergo a NLBB differed from lesions of patients who were send for a non-operative procedure. Furthermore, if a benign result was obtained, we assessed the total and kind of subsequent diagnostic procedures that were undertaken. A retrospective study on 718 women with 749 non-palpable breast lesions was performed. In 58% of women with non-palpable breast lesion, a non-surgical procedure was chosen. Lesions sent directly for surgery were more frequently not visible on ultrasound (62%) and mainly consisted of microcalcifications only (56%). In 45%, this primary surgical approach could have been avoided. If the non-operative procedure showed a non-malignant result, 41% of these women received an additional surgical diagnostic procedure. These figures obtained from routine daily practice show the importance of protocols in order to standardise diagnostic procedures and prevent unnecessary surgery.

Repeated mammographic screening reduces breast cancer mortality along the continuum of age.

OBJECTIVE: The optimal age boundaries for breast cancer screening are still under debate. A case-referent design was used to describe the effect of mammographic screening on breast cancer mortality along the continuum of age, based on a 20 year follow up period. SETTING: The population based breast cancer screening programme in Nijmegen, The Netherlands, which has biennially invited women over 35 years since 1975. METHODS: Cases, defined as women who died from primary breast cancer between 1987 and 1997, were selected from the group of women who received at least one invitation to the screening programme. For 157 cases, 785 women from the same group were selected as referents. Information on the index screening (the screening examination preceding diagnosis of the case) was collected for both cases and referents. The risk of dying from breast cancer was calculated per 10 year moving age group for women who had attended the index screening versus those who had not. RESULTS: The youngest 10 year age group showing an effect in our study were women aged 45-54 at their index screening. Breast cancer mortality for women in this group who attended the index screening was 32% lower, although not significant, than for women who did not (odds ratio (OR) 0.68, 95% confidence interval (95% CI) 0.33 to 1.41). This reduction in risk was not explained solely by an effect in women over 50 because the OR in women aged 45-49 was 0.56 (95% CI 0.20 to 1.61). Reductions in mortality became smaller with increasing age. Nevertheless, for women over 60 at index screening, participation in screening over a maximum 4 year period before diagnosis of the case yielded protective effects at least up to an age around 80. CONCLUSIONS: Although our results are based on a relatively small number of cases, they suggest that even in a programme with a 2 year screening interval there may be a benefit of starting screening around age 45. Also older women who participate at least once every 4 years still have much to gain from screening.

Reaction of carbon monoxide with the reduced active site of bacterial nitric oxide reductase.

Bacterial nitric oxide reductase (NOR), a member of the superfamily of heme-copper oxidases, catalyzes the two-electron reduction of nitric oxide to nitrous oxide. The key feature that distinguishes NOR from the typical heme-copper oxidases is the elemental composition of the dinuclear center, which contains non-heme iron (FeB) rather than copper (CuB). UV-vis electronic absorption and room-temperature magnetic circular dichroism (RT-MCD) spectroscopies showed that CO binds to Fe(II) heme b3 to yield a low-spin six-coordinate species. Photolysis of the Fe(II)-CO bond is followed by CO recombination (k(on) = 1.7 x 10(8) M(-1) x s(-1)) that is approximately 3 orders of magnitude faster than CO recombination to the active site of typical heme-copper oxidases (k(on) = 7 x 10(4) M(-1)x s(-1)). This rapid rate of CO recombination suggests an unimpeded pathway to the active site that may account for the enzyme's high affinity for substrate, essential for maintaining denitrification at low concentrations of NO. In contrast, the initial binding of CO to reduced heme b3 measured by stopped-flow spectroscopy is much slower (k(on) = 1.2 x 10(5) M(-1) x s(-1)). This suggests that an existing heme distal ligand (water/OH-) may be displaced to elicit the spin-state change observed in the RT-MCD spectrum.

Magnetic resonance imaging and mammography in women with a hereditary risk of breast cancer.

BACKGROUND: Although breast cancer screening is recommended to start at a younger age for women with a hereditary risk of breast cancer, the sensitivity of mammography for these women is reduced. We compared magnetic resonance imaging (MRI) with mammography to determine which is more sensitive and whether MRI could play a role in the early detection of breast cancer for these women. METHODS: We constructed a retrospective cohort of all breast MRI and mammography surveillance reports made in our department from November 1994 to February 2001. All of the 179 women in the cohort had received biannual palpation in addition to annual imaging by MRI, mammography, or both. The 258 MRI images and the 262 mammograms were classified with the use of the BI-RADS (i.e., Breast Imaging Reporting and Data System) scoring system, which has five categories to indicate the level of suspicion of a lesion. Receiver operator characteristic curves were generated for MRI and mammography, and the area under each curve (AUC) was assessed for the entire cohort of 179 women and for a subset of 75 women who had received both an MRI and a mammographic examination within a 4-month period. All statistical tests were two-sided. RESULTS: In the cohort of 179 women, we detected 13 breast cancers. Seven cancers were not revealed by mammography, but all were detected by MRI. For the entire cohort, the AUC for mammography was 0.74 (95% confidence interval [CI] = 0.68 to 0.79), and the AUC for MRI was 0.99 (95% CI = 0.98 to 1.0). For the subset of women who had both examinations, the AUC for mammography was 0.70 (95% CI = 0.60 to 0.80), and the AUC for MRI was 0.98 (95% CI = 0.95 to 1.0). CONCLUSION: MRI was more accurate than mammography in annual breast cancer surveillance of women with a hereditary risk of breast cancer. Larger prospective studies to examine the role of MRI in screening programs are justified.

Diverging breast cancer mortality rates in relation to screening? A comparison of Nijmegen to Arnhem and the Netherlands, 1969-1997.

Age-standardised breast cancer mortality rates have been stable for decades. However, rates have started to decline in several Western countries. In countries where population-based screening programmes for breast cancer were introduced in the late 1980s or early 1990s, the key question now is to what extent screening is responsible for the reported declines in mortality. This study compares breast cancer mortality rates in Nijmegen, where a screening programme for breast cancer was introduced in 1975, to a control city, Arnhem, and to the Netherlands as a whole over a 20-year period. Age-standardised breast cancer mortality rates as well as age-standardised mortality ratios were calculated for successive calendar years from 1969 to 1997. Further, a tailor-made period-cohort-group Poisson regression model was fitted. Figures displaying age-standardised mortality rates and ratios showed inconclusive patterns with regard to the expected impact of screening. Depending on when mortality rates were allowed to deviate between populations, the period-cohort-group analysis indicated a non-significant 6% to 16% reduction in breast cancer mortality after 2 decades in favour of the Nijmegen female population. Possible explanations are discussed as to why the mortality reductions reported by randomised trials might not be observed in a public health screening programme, such as the Nijmegen programme, evaluated by comparisons of geographical trends.

Breast carcinomas of limited extent: frequency, radiologic-pathologic characteristics, and surgical margin requirements.

BACKGROUND: Clinical trials established the value of breast-conserving treatment (BCT) including the macroscopic removal of the tumor followed by local radiation therapy (RT) for Stage I and II invasive carcinomas. The occurrence of local tumor recurrence is related to the extent and multifocality of the tumor. Various studies aim to identify those tumors that could be proper candidates for conventional BCT. Furthermore, recent studies have focused on the identification of tumors that may be treated by breast-conserving surgery alone without RT. Small, localized tumors theoretically should be the potential candidates for this type of treatment. The mammographic and pathologic criteria for the identification of tumors with limited extent are not yet established; furthermore, the optimal extent of the surgical excision and the method for margin examination are controversial. METHODS: Surgical breast-conserving procedures were simulated in a review of 135 mastectomy specimens of patients treated for an invasive carcinoma (> or = 4 cm in size, all pathologic types except invasive lobular carcinoma) who were theoretically eligible for conservative treatment. Tumor spread including possible multifocality and multicentricity was studied by the technique of correlated specimen radiography and pathology. Breast carcinoma of limited extent (BCLE), the proper tumor profile for BCT, was defined as having no invasive carcinoma, ductal carcinoma in situ, and lymphatic emboli foci beyond 1 cm from the edge of the dominant mass. RESULTS: Fifty-three percent of the patients in this series had a BCLE. No statistically significant relation was found between BCLE and patient age, pathologic size, type and grade of the tumor, lymph node status, mode of detection, and mammographic aspect of the index tumor. Based on mammography, the absence of calcification or tumor density beyond the edge of index tumor appears to be the best predictor for BCLE (P < 0.0001). A 1-cm microscopically tumor free margin as the outer rim of a macroscopic surgical margin of 2 cm gives the best positive predictive value based on pathology (P < 0.0001). By applying the above conditions, 72 of the 135 cancers were identified as being potential BCLE cases in this series. However, whereas 64 of these 72 tumors (89%) were correctly identified as being true BCLE, 8 (11%) were erroneously identified as such (non-BCLE cases), having "residual" tumor foci beyond 2 cm from the edge of the dominant tumor. CONCLUSIONS: We conclude, that approximately 50% of invasive ductal carcinomas may have limited extent. The accuracy of identifying this group of cancers, the proper candidates for BCT, by applying state-of-the-art mammography and pathology may be as high as 90%. A subset of these tumors might represent the potential candidates for treatment with surgery alone without RT. As a result, the routine application of BCT complemented by RT would have led to the overtreatment of 89% of the patients with a BCLE in this series; conversely, 11% of the tumors may have recurred without the use of RT. Considering that these conclusions are based on a theoretic morphologic model, further clinical studies with facilities for high quality team approach in diagnosis and therapy are needed to evaluate the impact of BCLE on BCT strategies. The results of this study should not justify the withholding of RT outside the context of clinical trials.

Nationwide breast cancer screening programme fully implemented in The Netherlands.

The period 1990-1997 saw the implementation of a nationwide breast cancer screening programme in the Netherlands, which provided biennial mammography for all women aged 50-69 years (50-75 years at present). The National Evaluation Team monitors the programme annually collecting regional data on screening outcomes; regional cancer registries provide data on interval cancers and on breast cancers in unscreened women by linkage of cancer registry data to data on screened women. Of 4 million women invited, 78.5% attended for screening. Screening resulted in 13.1 referrals, 9.2 biopsies and 6.1 breast cancers detected per 1000 women screened initially (6.9, 4.5 and 3.5 per 1000 in subsequently screened women, respectively). Within the first 2 years following screening 0.95 interval cancers per 1000 women-years were diagnosed. The stage distribution of screen-detected cancers was more favourable than that of interval cancers and of those diagnosed in unscreened women. The results are largely consistent with expectations. Results may nonetheless be further improved, particularly the detection rate in subsequent screens.

Automated classification of clustered microcalcifications into malignant and benign types.

The objectives in this study were to design and test a fully automated method for classification of microcalcification clusters into malignant and benign types, and to compare the method's performance with that of radiologists. A novel aspect of the approach is that the relative location and orientation of clusters inside the breast was taken into account for feature calculation. Furthermore, correspondence of location of clusters in mediolateral oblique (MLO) and cranio-caudal (CC) views, was used in feature calculation and in final classification. Initially, microcalcifications were automatically detected by using a statistical method based on Bayesian techniques and a Markov random field model. To determine malignancy or benignancy of a cluster, a method based on two classification steps was developed. In the first step, classification of clusters was performed and in the second step a patient based classification was done. A total of 16 features was used in the study. To identify meaningful features, a feature selection was applied, using the area under the receiver operating characteristic (ROC) curve (Az value) as a criterion. For classification the k-nearest-neighbor method was used in a leave-one-patient-out procedure. A database of 192 mammograms with 280 true positive detected microcalcification clusters was used for evaluation of the method. The set consisted of cases that were selected for diagnostic work up during a 4 year period of screening in the Nijmegen region (The Netherlands). Because of the high positive predictive value in the screening program (50%), this set did not contain obvious benign cases. The method's best patient-based performance on this set corresponded with Az = 0.83, using nine features. A subset of the data set, containing mammograms from 90 patients, was used for comparing the computer results to radiologists' performance. Ten radiologists read these cases on a light-box and assessed the probability of malignancy for each patient. All participants had experience in clinical mammography and participated in our observer study during the last 2 days of a 2-week training session leading to screening mammography certification. Results on the subset showed that the method's performance (Az = 0.83) was considerably higher than that of the radiologists (Az = 0.63).

An automatic method to discriminate malignant masses from normal tissue in digital mammograms.

Specificity levels of automatic mass detection methods in mammography are generally rather low, because suspicious looking normal tissue is often hard to discriminate from real malignant masses. In this work a number of features were defined that are related to image characteristics that radiologists use to discriminate real lesions from normal tissue. An artificial neural network was used to map the computed features to a measure of suspiciousness for each region that was found suspicious by a mass detection method. Two data sets were used to test the method. The first set of 72 malignant cases (132 films) was a consecutive series taken from the Nijmegen screening programme, 208 normal films were added to improve the estimation of the specificity of the method. The second set was part of the new DDSM data set from the University of South Florida. A total of 193 cases (772 films) with 372 annotated malignancies was used. The measure of suspiciousness that was computed using the image characteristics was successful in discriminating tumours from false positive detections. Approximately 75% of all cancers were detected in at least one view at a specificity level of 0.1 false positive per image.

[Alterations of stage distribution for breast cancer since the implementation of national screening program in the Netherlands during 1989-1995]. / Veranderde stadiumverdeling van borstkanker sinds de invoering van de landelijke screening; onderzoek over de periode 1989-1995.

OBJECTIVE: To determine the results of the nation-wide breast cancer screening on the basis of the stage distribution of diagnosed breast cancer 1989-1995. DESIGN: Descriptive, inventory. METHOD: At the University of Nijmegen data concerning the numbers of diagnosed breast cancer cases and age and stage at diagnosis were obtained from the Netherlands Cancer Registry for the years 1989-1995. Population data were provided by Statistics Netherlands. With these data incidence figures were calculated. RESULTS: The number of invasive breast cancers rose from almost 8,000 in 1989 to almost 10,500 in 1994. In 1995, a small decrease was visible. The increase was largest in women aged 50-69 years, the target population of the screening programme (introduced in 1988), and less in older women. In women aged 50-69 years, the incidence of stage I tumours doubled, stage II tumours increased but had returned to original levels in 1995, whereas stage III and IV had decreased by more than 24%. The number of carcinomas in situ rose from 336 in 1989 to 829 in 1995 in all age categories, but most in women aged 50-69 years. CONCLUSION: When the implementation scheme of the national screening programme is taken into account, the results are indicative of a reduction in breast cancer mortality. Whether such a mortality reduction actually will follow, has to be awaited.

Parity and mammographic breast density in relation to breast cancer risk: indication of interaction.

We examined whether the harmful influence of nulliparity on breast cancer risk could be mediated by high mammographic density. Another possibility is that mammographic density and nulliparity act independently or perhaps synergistically on breast cancer risk. Our study population consisted of 129 cases and 517 controls who had been participants in the Nijmegen breast cancer screening programme for 10 years. Breast density was classified with a fully automated technique on digitized mammograms from the screening examination 10 years before diagnosis. Classification was based on the proportion of the breast that was composed of high density: < 5%, 5-25% or > 25%. Data on parity and potential confounders were obtained using a questionnaire, administered at the same examination. We found that nulliparae with low breast density (< 5%) were not at increased risk compared to parous women with low density: OR 1.1 (95% CI 0.2-5.8). Parous women with < 5% density formed the reference category throughout all analyses. The risks for parous women with 5-25% or > 25% density were 2.7 (95% CI 1.3-5.6) and 3.6 (95% CI 1.7-7.7) fold increased, respectively. However, when both factors were present (nulliparity and > or = 5% density), breast cancer risk was 7.1 times higher (95% CI 3.2-15.9). This could indicate that nulliparity and high breast density might work synergistically and that breast density is not just an explanatory factor in the influence of nulliparity on breast cancer risk. It is hypothesized that high breast density (reflecting fibro-glandular tissue with increased epithelial cell proliferation) is more susceptible to carcinogenic effects in the undifferentiated epithelial breast tissue of nulliparae than in the differentiated tissue of parous women. Since there were few data, no firm conclusions can be drawn. If these findings can be confirmed in a larger study population, however, they may have important implications for the prevention and early detection of breast cancer.

[Reading screening mammograms with the help of neural networks]. / Beoordeling van screeningsmammogrammen met behulp van neurale netwerken.

With digital mammography it is possible to assist radiologists in breast cancer screening with computers to improve their reading performance. The need for this has been demonstrated by studies showing a large variability in skill of radiologists reading mammograms. Moreover, retrospective studies show that a significant number of cancers are clearly visible on earlier screening mammograms, even for 'trained' computers. Methods for automated detection of breast cancer in mammograms often use artificial neural networks. These are 'trained' to recognize abnormal mammographic areas using a large database of known cases. For detection of microcalcification clusters very reliable algorithms exist, with such high sensitivity that radiologists can limit their search to areas that have been marked 'suspect' by the computer. The development of methods to recognize malignant masses is much more difficult, but ample progress has been achieved in recent years.