Among cancer screenings targeted at specific organs, breast cancer screening is the one that has been evaluated most critically in terms of its effectiveness in reducing mortality, on the basis of randomized controlled trial (RCT) or case control studies. The standard modality for breast cancer screening worldwide is mammography. The data collected by the International Breast Cancer Screening Network were summarized in comparison with the current situation in Japan. Most countries now perform screening mammography for women aged 50 years and over, in view of the age distribution of breast cancer incidence. The highest incidence in western countries is in the seventh or eighth decade of life, whereas that in Japan is in the fifth decade. Therefore the introduction of mammography for women aged 40-49 has become an important issue in Japan. More efforts to improve quality control, not only in facilities but also the qualifications of personnel, should be made to reduce mortality due to breast cancer.

View full abstract

Screening mammography for breast cancer examinations has been carried out in Yokohama City since October in 2001. 18, 417 eligible women were screened, 12.6% required more detailed examination and 0.8% were diagnosed with cancer. 17 cases of breast cancer were found among the examinees that were not found any problems by physical examination, showing the efficacy of screening with mammography. 20.8% of breast cancers found on this program were detected by double-checking system on mammogram reading. The quality control of the screening system has been effective to reduce false-positive rate on this program.

View full abstract

We ceased to use ultrasonography in Toyama Prefecture in April 2001, and instead adopted mammography for breast cancer screening. The current program covers all areas in our prefecture including 35 cities, towns and villages. We have 5 mammography machines, which are carried by 3 cars. Breast cancer screening is performed for women aged 30 years or older, and mammography is routinely carried out for all examinees aged 40 years or older. Here the results are reviewed and newly encounted problems are discussed.
Although the number of participants had decreased previously, it rose to 36, 278 in 2001 and to 39, 122 in 2002. The results of breast cancer screening in Toyama Prefecture in 2001 and 2002 were as follows. The recall rate was 5.9% and 4.8%, the number of women whose breast cancer was detected for the first time was 88 and 75, and the breast cancer detection rate among all participants was 0.24% and 0.19%, respectively. The cancer detection rate in women aged 50 years or older was 0.26% and that in women in their 5 th decade was 0.17%, these figures being higher than that of 0.08% obtained in previous years when mass screening had been carried out by ultrasonography. The ratio of early-stage cancers was 67. 3% at the recent mammography screening, which was superior to the figure of 62. 8% obtained at the previous ultrasonography screening. In addition, the rate of detection of non-invasive cancers increased markedly following the introduction of mammography. All these results suggest that mammography is superior for breast cancer screening.
It is noteworthy that, although mammography was introduced in 2001, 18 women had cancers that were missed at screening in 2001 but were detected in 2002. The false negative result in 2001 was due to technical failure by radiographers, misinterpretation by doctors, or both. Our mammography reading has been carried out using a double-check method by 20 pairs of doctors. Each individual pair of doctors produced fairly different results : the recall rate ranged from 2.7% to 24.6% and the rate of correct diagnosis of breast cancer ranged from 0.6% to 13.3%. We asked almost all radiography technicians and doctors who had participated in mammography screening to attend a review meeting twice a year. The meetings were effective for improving the quality of both mammograms and correct reading.

View full abstract

Mammography was introduced for breast cancer screening in Fukui Prefecture in November, 2001. Breast cancer was detected by this method in 11 (0.48%) of 2, 314 women examined in 2002. The cancer detection rate was satisfactory, but the recall rate was as high as 16.1%. Therefore we recognized the need to decrease the recall rate by improving the quality control of mammography.
We have adopted a separate A-type screening in which mammograms are obtained in a cancer screening car and the films are read twice by different doctors, the first reading being performed by 16 less-experienced doctors and the second reading by 6 experienced doctors, each independently. The high recall rate might have arisen for three reasons : (1) the number of mammograms read by one doctor was not enough to promote reading ability, (2) the first reader was not acquainted with the second reader's interpretation, and (3) there was no chance to discuss specific cases in detail, and there was no training course in mammogram reading.
In order to solve these problems, (1) we set up a training course in mammogram reading, (2) devised a system for acquainting the first reader with the second reader's interpretation, and (3) made all the readers' interpretations available. As a result, the recall rate dropped to 11.5%. This has encouraged us to continue such trials to improve the quality control of mammography.

View full abstract

There are two major problems with breast cancer screening that require urgent resolution : improvement of quality control and an increase in the participation rate.
For breast cancer screening by physical examination including inspection and palpation combined with mammography, quality control is required for each of the two types of examination. For improvement of mammography quality control, many clinicians throughout Japan have attended training courses carried out under the auspices of the Central Committee on Quality Control of Mammographic Screening and been authorized by passing the post-course test. On the other hand, the quality control of physical examination has been overlooked.
Starting in 2000, Okayama Prefectural Medical Association held three successive training courses, with a total of 6 hours, with the aim of promoting the quality of inspection and palpation at breast cancer screening, and so far over 1, 000 physicians have attended. Those who had mastered the basic requirement of physical examination began to conduct breast cancer screening in 2001. As a result, the cancer detection rate rose from 0.05% to 0.17% in Okayama City, and the rate throughout Okayama Prefecture as a whole rose from 0.05% to 0.13%. Therefore we think that training of physicians in inspection and palpation is important.
In order to increase the participation rate, it will be necessary for the Prefectural Medical Association to set up an authorized system to call on all primary-care physicians to participate in the screening.

View full abstract

[Purpose] To evaluate quantitatively the efficacy of biennial mammographic screening of breast cancer in Japanese women aged 40-49 years in terms of cause-specific mortality reduction.
[Subjects and Methods] Three identical groups were assumed to be present, each group consisting of 100, 000 Japanese women aged 40-49 years. The first group underwent biennial mammographic screening, the second group underwent annual mammographic screening, and the third group had no screening. A deterministic mathematical model for cancer screening developed by the author was applied to calculate the number of deaths due to breast cancer among the three groups. Thus the relative risk (RR) and the risk difference (RD) of death were obtained between the screened group and the unscreened group in order to show the efficacy of the screening.
[Results] The RR of the biennial screened group/unscreened group was 0.66 and that of the annual screened group was 0.57. The RD of the two screened groups/unscreened group were 10. 6 and 14. 8 persons/year. It can be seen that annual screening was more effective than biennial screening, although the latter still retained good values of RR and RD.
[Conclusion] Biennial mammograpic screening for Japanese women aged 40-49 years may present a RR of 0.66 compared to unscreened women, although it is worse than that of annual screening. Cost-effectiveness analysis of biennial screening will probably show that it is superior to annual screening.

View full abstract

When breast cancer screening is performed, quality control is essential. We investigated the utility of cancer detection rate as an indicator of screening outcome. The subjects were 37, 758 women aged 50-69 yr examined at Miyagi Cancer Association from 1999 to 2002. We surveyed the screening results according to screening history and compared them with morbidity in the Miyagi cancer registry. The breast cancer detection rate at first-time mammographic examinations was O. 31% and that at biennial mammographic examinations was 0.17%. The rate for expected morbidity was 94.5%, which was almost the same as the sensitivity (95.0%) of this screening which we reported previously. It is possible to estimate the sensitivity of screening from the cancer detection rate of biennial examinations. It is important to survey screening results according to screening history in order to evaluate the accuracy of screening.

View full abstract

We have reported on the quality control of training courses in mammogram interpretation. Here we consider the planning of the courses and attitude of the participants based on analysis of 26 courses held between January 2002 (Akita 1) and March 2003 (Akita 2). The same questionnaire method, ranking and quality clearance as those in the previous reports were employed. Quality clearance was 64% (7 among 11) for alpha-ranked courses and 67% (10 among 15) for beta-ranked courses (there was no gamma-ranked course during this period), or 65% in all, which was much worse than the figure of 82% reported previously. As in the previous study, there was a tendency for those who had prepared themselves to show better film-reading ability. Also, doctors with radiology technicians in their facilities who had already taken a technical training course to produce mammograms of high quality tended to learn technical terms and category classifications more easily and earned higher grades in the post-course test. One notable finding was an increase in the number of comments stating that certain lectures that were not helpful for passing the test were unnecessary.These film-reading tests are intended to improve the accuracy of mammography reading, which is indispensable for maintaining high-quality mammographic breast cancer screening, and are not merely for passing tests. The planners of the courses who seek for the joint auspices of the Central Committee for Quality Control of Mammographic Screening (CCQCMS) need to recognize this, inform participants about it, and make efforts to hold small group training sessions and workshops before and after the training courses. The attending doctors are required to prepare themselves at such pre-course workshops and to keep their film-reading ability at the post-course level.

View full abstract

Images of a contrast-detail phantom made with polymerized methyl methacrylate were acquired using screen-film (MinR2000/MinR2000) and full-field digital mammography systems with the same exposure conditions. Observer performance tests were conducted with participating eight observers. Each observer was required to identify a randomly located disk in phantom images with a 50% confidence level. Average contrast-detail curves for the digital and film images were obtained from the observation results. Image quality figures (IQF) were also calculated from the individual observer performance tests. The results indicated that the digital contrast-detailed curves and IQF are, on average, superior to those of the screen-film system.

View full abstract

Among 87 breast cancer lesions found during clinical breast cancer screening with combined use of mammography (MMG) and ultrasonography (US), we examined 20 lesions that were not detected by MMG and 13 lesions that were not detected by US. Nine of the 20 lesions initially undetected by MMG were detected by closer examination. However, 3 of these 9 lesions were detected by spot MMG only, it was thought that detection by screeing MMG would be difficult. We considered that the remaining 6 lesions could have been detected by screening MMG if more care had been taken with regard to positioning or sufficient pressure. Twelve of the 13 lesions initially undetected by US were subjected to closer examination, and 7 of them were detected. However, 5 of these 7 lesions were detected as echogenic spots by scanning after it had become clear they were abnormal. Histologically they were non-invasive or partially invasive ductal carcinoma and papillotubular carcinoma, and considered difficult to detect by screening US. The remaining 2 lesions were invasive ductal carcinomas which were detectable as mass lesions by closer US examination, and these were considered detectable by screening US.
In summary, it was impossible to detect 14 lesions (16.1%) by MMG and 10 lesions (11.5%) by US among 87 lesions of breast cancer found during the course of breast cancer screening.

View full abstract

Hand-scanning ultrasound mass-screening (hereafter referred to as US screening) over the past 15 years has indicated a higher cumulative survival rate in the mass screening group than in the outpatient group in women aged both under 50 and 50 or over. The accuracy and capacity of US screening and the proportion of patients with subjective symptoms were studied for each group. We discovered that women under 50 had a significantly higher rate of “further examination needed” and “having subjective symptoms”, and a significantly lower rate of “positive predictive value” than women aged 50 or over. There was no difference in the detection rate, sensitivity or specificity between the two groups.
We verified that the mass screening group had a significantly higher ratio of non-palpable breast cancer, a significantly smaller tumor caliber and a significantly earlier stage in terms of n-factor and tnm-classification than the outpatient group. Although pathological investigation indicated significantly more non-invasive breast cancers in the mass screening group than in the outpatient group, we also confirmed that there was no significant difference between the two groups in the ratio of invasive ductal cancer and non-invasive cancer subtype classification. These results suggest that US screening indicates a better survival rate because it detects breast cancers earlier, not because it detects only those with lower malignancy.
We verified that 26.6% of non-palpable breast cancers were detected only by US. Furthermore, we found that in women under 50, because of breast stiffness, the ratio of non-palpable breast cancers was 30.2%, which was larger than in the older age group, and of these, 69.2% were diagnosed as invasive ductal breast cancer. There is a risk that mass-screening by physical examination alone will overlook these cases, and therefore we advocate the introduction of a safe and useful US screening program for women under 50 who are currently examined by palpation only by the Japanese Ministry of Health and Welfare.

View full abstract