| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Commentary |
1 Department of Radiology, Box 1234, The Mount Sinai School of Medicine, 1 Gustave L. Levy PI., New York, NY 10029-6574.
Received November 6, 2001;
accepted after revision November 6, 2001.
This article is a commentary on the preceding article by Taplin et al.
Introduction
 Top Introduction Reduction in Breast Cancer...Clinical Outcomes Attributed to...Clinical Outcomes Resulting from...ConclusionReferences |
|---|
Results from screening trials conducted in Europe indicate that even greater benefits may be expected as more American women comply with recent guidelines advising annual mammography, beginning when they are 40 years old [4,5,6]. Because better image quality should lead to earlier detection and further reduction in breast cancer deaths, we must strive for the best possible images obtainable from current techniques while continuing to research improved imaging methods.
Reduction in Breast Cancer Death Rates Through Screening
|
TopIntroductionReduction in Breast Cancer...Clinical Outcomes Attributed to...Clinical Outcomes Resulting from...ConclusionReferences |
|---|
Before the Swedish service screening study, proof of benefit from screening mammography had been obtained from randomized clinical trials. These trials compared breast cancer death rates in study groups who were offered screening and control groups who were not offered screening during the same time period. One randomized clinical trial, conducted by Tabar et al. [8] in the same two Swedish counties from 1978 to 1985, found a 30% decrease in breast cancer deaths among women who were 40-74 years old at entry into screening. These results are impressive, but for several reasons randomized clinical trials underestimate the maximum benefit that may be derived from screening. For example, some women in the study group may decline to be screened and some women in the control group may obtain screening on their own outside the trial. The amount of benefit is also limited by the technical quality of mammography performed during the trial.
Nine randomized clinical trials of screening mammography have been conducted: the Health Insurance Plan of Greater New York trial [9]; five Swedish trials [8, 10,11,12,13,14]; the Edinburgh, Scotland, trial [15, 16]; and the two arms of the Canadian National Breast Screening Study [17, 18]. Among these trials, outcomes have varied. Reports of breast cancer mortality reduction have ranged from none to 45%. Variation in benefits has been attributed to differences among these trials in study design, execution, and technical quality of mammography. In general, results have been sufficiently favorable to enable most major medical organizations in the United States, including the American Cancer Society, the American Medical Association, and the American College of Radiology, to advise women to undergo annual screening mammography, beginning when they are 40 years old [4,5,6].
Clinical Outcomes Attributed to Image Quality: Interstudy
Comparisons
|
TopIntroductionReduction in Breast Cancer...Clinical Outcomes Attributed to...Clinical Outcomes Resulting from...ConclusionReferences |
|---|
Detection of minimal carcinoma, defined as every in situ carcinoma regardless of size and any invasive cancer smaller than 1 cm, represents another difference in outcome that has been related to better mammographic technique at the Breast Cancer Detection Demonstration Project. There, about 25% of detected cancers were classified as minimal, a fourfold increase from the 8% rate at the Health Insurance Plan of Greater New York trial [19, 20].
The Swedish Two-County Trial (1977-1985) used modern breast compression techniques and extended processing—innovations that were not available during the Health Insurance Plan of Greater New York trial (1963-1969). Partly as a result of improved mammographic technique, the Swedish Two-County Trial showed a greater reduction in breast cancer mortality (30% vs 24%), even though only the Health Insurance Plan of Greater New York study included clinical examination and used more frequent screening intervals (1 year vs 2-3 years) and a greater number of standard mammographic views (two vs one) [8,9,10].
The Kopparberg arm of the Swedish Two-County trial (1977-1985) used single mediolateral oblique view mammography with extended processing but did not use mammographic grids. A subsequent service screening program in Kopparberg (1989-1992) used two standard mammographic views (mediolateral oblique and craniocaudal), extended processing, and a moving grid. Although the predicted survival rate for women 40-49 years old with invasive carcinoma in the Kopparberg series was not significantly different from that for women in the trial (83% vs 82%), a fourfold increase in ductal carcinoma in situ was found among cancers detected at the first screening (39% vs 11%) [10].
During the course of the screening trial in Edinburgh, Scotland (1979-1987), there was a progressive increase in screen sensitivity (screen-detected cancers / screen-detected and interval cancers) from 92% to 97%. There was also a decrease in the proportional incidence of interval cancers (interval cancer rates x 100 / control group incidence) from 28% to 5%. The investigators attributed these results to continuous improvement in their mammographic technique as the trial progressed [15].
Deficiencies in the technical quality of mammography at the Canadian National Breast Screening Study (1980-1990) have been well documented by means of several independent external evaluations [21, 22]. In one review, 67% of mammograms obtained during the year 1985 were judged as technically unacceptable [21]. Poor technical quality has been used to explain why the Canadian National Breast Screening Study-1 was unable to demonstrate any mortality rate reduction from mammographic screening of women 40-49 years old [17, 21, 22]. Poor technical quality also indicates why the Canadian National Breast Screening Study-2 found no evidence of additional benefit from screening women 50-59 years old by means of both mammography and physical examination combined, compared with screening by physical examination alone [18].
Because of the technical advances that have been made since randomized clinical trials were performed in the 1980s, results of those trials underestimate the potential benefit of screening mammography, according to Sickles and Kopans [23]. They found that advanced cancers represented a substantially lower proportion of malignancies at service screening programs in British Columbia and at the University of California, San Francisco, in the early 1990s than at the randomized clinical trials conducted in earlier decades. Findings for the later screenings compared with the randomized clinical trials were, respectively, 12-14% versus 22-30% for T2 or larger lesions, 12-13% versus 29-43% for tumors having findings positive for malignancy in the axillary lymph nodes, and 15-19% versus 24-58% for malignancies classified as stage II or higher [23].
Clinical Outcomes Resulting from Image Quality: Intrastudy
Comparisons
|
TopIntroductionReduction in Breast Cancer...Clinical Outcomes Attributed to...Clinical Outcomes Resulting from...ConclusionReferences |
|---|
In a study conducted at 31 screening centers in the United Kingdom National
Health Service Breast Screening Programme, Young et al.
[24] compared the detection
rate for small invasive cancers (greatest diameter, 
10 mm) with the
average film density at each center. The detection rate was 0.12% at centers
where film densities were less than 1.2 density units and the detection rate
was 0.17% at centers using higher film densities. The authors concluded that
raising film density to 1.2 or higher might increase the detection rate of
small invasive cancers by as much as 50%.
As a result of this study, the United Kingdom National Health Service Breast Screening Programme recommended use of a target film density in the range of 1.4-1.8 to ensure adequate exposure [25]. The American College of Radiology Mammography Accreditation Program has a similar recommendation for an optical density of 1.4-2.0 for mammography [26]. These recommendations are also based on phantom studies that have found higher test object scores at these higher density levels [27].
In a subsequent study, Young et al. [25] showed that the beneficial effect of using a high optical density was most marked at facilities using two mammographic views (craniocaudal and mediolateral oblique) per breast. Conversely, the beneficial effect of using two-view mammography was greater at facilities using a density of at least 1.4. Relative to a base level of one view and a film density of less than 1.4, a highly significant 25% increase in detection of invasive cancers was observed at facilities using two views, providing they used higher optical densities.
The accompanying article by Taplin et al. establishes another direct link between mammographic image quality and clinical outcome. The authors found that patients with either improper breast positioning or poor overall image quality had subsequently higher interval cancer rates.
|
TopIntroductionReduction in Breast Cancer...Clinical Outcomes Attributed to...Clinical Outcomes Resulting from...ConclusionReferences |
|---|
Studies such as the one reported by Taplin et al. in this issue confirm the practical value of the efforts of the American College of Radiology and the United States Food and Drug Administration. Their study should encourage efforts to further ensure proper image quality in each of our own practices and around our nation. Their publication should stimulate the initiation of additional evidence-based studies to document that breast positioning, compression, exposure, contrast, sharpness, and noise can each influence clinical outcomes.
|
TopIntroductionReduction in Breast Cancer...Clinical Outcomes Attributed to...Clinical Outcomes Resulting from...ConclusionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  L. Berlin Breast Cancer, Mammography, and Malpractice Litigation: The Controversies Continue Am. J. Roentgenol., May 1, 2003; 180(5): 1229 - 1237. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
