Per-Henrik Zahl

The Natural History of Invasive Breast Cancers Detected by Screening Mammography

BACKGROUND The introduction of screening mammography has been associated with sustained increases in breast cancer incidence. The natural history of these screen-detected cancers is not well understood. METHODS We compared cumulative breast cancer incidence in age-matched cohorts of women residing in 4 Norwegian counties before and after the initiation of biennial mammography. The screened group included all women who were invited for all 3 rounds of screening during the period 1996 through 2001 (age range in 1996, 50-64 years). The control group included all women who would have been invited for screening had there been a screening program during the period 1992 through 1997 (age range in 1992, 50-64 years). All women in the control group were invited to undergo a 1-time prevalence screen at the end of their observation period. Screening attendance was similar in both groups (screened, 78.3%, and controls, 79.5%). Counts of incident invasive breast cancers were obtained from the Norwegian Cancer Registry (in situ cancers were excluded). RESULTS As expected, before the age-matched controls were invited to be screened at the end of their observation period, the cumulative incidence of invasive breast cancer was significantly higher in the screened group than in the controls (4-year cumulative incidence: 1268 vs 810 per 100 000 population; relative rate, 1.57; 95% confidence interval, 1.44-1.70). Even after prevalence screening in controls, however, the cumulative incidence of invasive breast cancer remained 22% higher in the screened group (6-year cumulative incidence: 1909 vs 1564 per 100 000 population; relative rate, 1.22; 95% confidence interval, 1.16-1.30). Higher incidence was observed in screened women at each year of age. CONCLUSIONS Because the cumulative incidence among controls never reached that of the screened group, it appears that some breast cancers detected by repeated mammographic screening would not persist to be detectable by a single mammogram at the end of 6 years. This raises the possibility that the natural course of some screen-detected invasive breast cancers is to spontaneously regress.

Breast cancer mortality in organised mammography screening in Denmark: comparative study

Objective To determine whether the previously observed 25% reduction in breast cancer mortality in Copenhagen following the introduction of mammography screening was indeed due to screening, by using an additional screening region and five years additional follow-up. Design We used Poisson regression analyses adjusted for changes in age distribution to compare the annual percentage change in breast cancer mortality in areas where screening was used with the change in areas where it was not used during 10 years before screening was introduced and for 10 years after screening was in practice (starting five years after introduction of screening). Setting Copenhagen, where mammography screening started in 1991, and Funen county, where screening was introduced in 1993. The rest of Denmark (about 80% of the population) served as an unscreened control group. Participants All Danish women recorded in the Cause of Death Register and Statistics Denmark for 1971-2006. Main outcome measure Annual percentage change in breast cancer mortality in regions offering mammography screening and those not offering screening. Results In women who could benefit from screening (ages 55-74 years), we found a mortality decline of 1% per year in the screening areas (relative risk (RR) 0.99, 95% confidence interval (CI) 0.96 to 1.01) during the 10 year period when screening could have had an effect (1997-2006). In women of the same age in the non-screening areas, there was a decline of 2% in mortality per year (RR 0.98, 95% CI 0.97 to 0.99) in the same 10 year period. In women who were too young to benefit from screening (ages 35-55 years), breast cancer mortality during 1997-2006 declined 5% per year (RR 0.95, CI 0.92 to 0.98) in the screened areas and 6% per year (RR 0.94, CI 0.92 to 0.95) in the non-screened areas. For the older age groups (75-84 years), there was little change in breast cancer mortality over time in both screened and non-screened areas. Trends were less clear during the 10 year period before screening was introduced, with a possible increase in mortality in women aged less than 75 years in the non-screened regions. Conclusions We were unable to find an effect of the Danish screening programme on breast cancer mortality. The reductions in breast cancer mortality we observed in screening regions were similar or less than those in non-screened areas and in age groups too young to benefit from screening, and are more likely explained by changes in risk factors and improved treatment than by screening mammography.

Natural history of breast cancers detected in the Swedish mammography screening programme: a cohort study

BACKGROUND The natural history of screen-detected breast cancers is not well understood. A previous analysis of the incidence change during the introduction of the Norwegian screening programme in the late 1990s suggested that the natural history of many screen-detected invasive breast cancers is to regress spontaneously but the study was possibly confounded by use of hormone replacement therapy in the population. We did a similar analysis of data collected during an earlier period when few women were exposed to hormone replacement therapy. METHODS We compared cumulative breast cancer incidence in age-matched cohorts of women living in seven Swedish counties before and after the initiation of public mammography screening between 1986 and 1990. Women aged 40-49 years were invited to screening every year and women aged 50-74 years were invited every 2 years. A screened group including all women aged 40-69 years (n=328,927) was followed-up for 6 years after the first invitation to the programme. A control group including all women in the same age range (n=317,404) was also followed-up for 6 years--4 years without screening and 2 years when they entered the screening programme. Screening attendance was much the same in both groups (close to 80%). Counts of incident invasive breast cancers were obtained from the Swedish Cancer Registry (in-situ cancers were excluded). FINDINGS Before the age-matched controls were invited to be screened at the end of their follow-up period, the 4-year cumulative incidence of invasive breast cancer was significantly higher in the screened group (982 per 100,000) than it was in the control group (658 per 100,000) (relative risk [RR] 1·49, 95% CI 1·41-1·58). Even after prevalence screening in the control group, the screened group had higher 6-year cumulative incidence of invasive breast cancer (1443 per 100,000 vs 1269 per 100,000; RR 1·14, 1·10-1·18). INTERPRETATION Because the cumulative incidence among controls did not reach that of the screened group, we believe that many invasive breast cancers detected by repeated mammography screening do not persist to be detected by screening at the end of 6 years, suggesting that the natural course of many of the screen-detected invasive breast cancers is to spontaneously regress. FUNDING None.

Overestimated lead times in cancer screening has led to substantial underestimation of overdiagnosis.

Background:Published lead time estimates in breast cancer screening vary from 1 to 7 years and the percentages of overdiagnosis vary from 0 to 75%. The differences are usually explained as random variations. We study how much can be explained by using different definitions and methods.Methods:We estimated the clinically relevant lead time based on the observed incidence reduction after attending the last screening round in the Norwegian mammography screening programme. We compared this estimate with estimates based on models that do not take overdiagnosis into account (model-based lead times), for varying levels of overdiagnosis. Finally, we calculated overdiagnosis adjusted for clinical and model-based lead times and compared results.Results:Clinical lead time was about one year based on the reduction in incidence in women previously offered screening. When overdiagnosed tumours were included, the estimates increased to 4–9 years, depending on the age at which screening begins and the level of overdiagnosis. Including all breast cancers detected in women long after the end of the screening programme dilutes the level of overdiagnosis by a factor of 2–3.Conclusion:When overdiagnosis is not taken into account, lead time is substantially overestimated. Overdiagnosis adjusted for model-based lead time is a function tending to zero, with no simple interpretation. Furthermore, the estimates are not in general comparable, because they depend on both the duration of screening and duration of follow-up. In contrast, overdiagnosis adjusted for clinically relevant tumours is a point estimate (and interpreted as percentage), which we find is the most reasonable method.

Why mammography screening has not lived up to expectations from the randomised trials

We analysed the relation between tumour sizes and stages and the reported effects on breast cancer mortality with and without screening in trials and observational studies. The average tumour sizes in all the trials suggest only a 12% reduction in breast cancer mortality, which agrees with the 10% reported in the most reliable trials. Recent studies of tumour sizes and tumour stages show that screening has not lowered the rate of advanced cancers. In agreement with this, recent observational studies of breast cancer mortality have failed to find an effect of screening. In contrast, screening leads to serious harms in healthy women through overdiagnosis with subsequent overtreatment and false-positive mammograms. We suggest that the rationale for breast screening be urgently reassessed by policy-makers. The observed decline in breast cancer mortality in many countries seems to be caused by improved adjuvant therapy and breast cancer awareness, not screening. We also believe it is more important to reduce the incidence of cancer than to detect it ‘early.’ Avoiding getting screening mammograms reduces the risk of becoming a breast cancer patient by one-third.

Overdiagnosis of breast cancer after 14 years of mammography screening.

BACKGROUND In 2004 we wrote in Tidsskriftet that mammography screening resulted in massive over-diagnosis and over-treatment of breast cancer. Our study was criticised because we had only five years of follow-up time and did not take account of the fact that increased use of hormone replacement therapy could lead to more breast cancer. We have now been screening women for 14 years, and during a period when the use of hormones has fallen by 70 %. MATERIAL AND METHOD Age-specific incidence rates, detection rates and interval rates for breast cancer in the period 1991-2009 have been computed for 40-79 year-old women. Incidence trends have been calculated using Poisson regression. RESULTS The incidence of breast cancer in the age group 40-49 was stable throughout the period, but rose by 50 % in the age group 50-69 years immediately after the start of screening. There was no significant reduction in the incidence of breast cancer in the age group 70-74. The number of new cases of breast cancer in the period increased from around 2000 to 2750. About 300 cases of ductal carcinoma in situ (DCIS) were also diagnosed. Today a total of some 1050 more women have been diagnosed than before screening started. Our calculations indicate that in the absence of screening, around 800 of these women would never have become breast cancer patients. INTERPRETATION The figures from 14 years of mammography screening indicate that all increase in the incidence of breast cancer is due to over-diagnosis: findings of tumours that in the absence of screening would never have given rise to clinical illness.

Breast Cancer Screening in Denmark: A Cohort Study of Tumor Size and Overdiagnosis

Effective breast cancer screening should reduce the incidence of advanced tumors (1). Tumors larger than 20 mm at detection are usually considered advanced because they are equivalent to T2 or greater in the TNM (tumor, node, metastasis) classification system (24). Screening mammography detects many small tumors that would not have become clinically evident in the remaining lifetime without screening (overdiagnosis) (5). Whether screening reduces the incidence of advanced tumors has important therapeutic implications. Overdiagnosed lesions may be unnecessarily treated with surgery, chemotherapy, and radiation, which subjects women to the harms of therapy without benefit (6). Our study objectives were to examine the association of screening with a reduction in the incidence of advanced cancer and estimate the level of overdiagnosis in the Danish breast screening program. Denmark provides a unique opportunity to estimate overdiagnosis because only 20% of the population aged 50 to 69 years was invited to participate in a mammography screening program for 17 years, which, to our knowledge, is the longest available period with differential mammography access in any country. Unlike studies in other settings that did not have a contemporaneous nonscreened group or nonscreened age groups (3, 4, 7, 8), we examine the association of screening with incidence of advanced cancer and estimate overdiagnosis using the contemporaneous same-age nonscreened group and nonscreened age groups as controls. Methods Design We conducted a cohort study using individual, anonymized data on tumor size for all Danish women aged 35 to 85 years diagnosed with invasive breast cancer during 1980 to 2010 from the Danish Breast Cancer Group (DBCG) and the Danish Cancer Registry (DCR). These independent databases cover all of Denmark and have been validated; about 1.2% of cases of breast cancer were missing or misdiagnosed (9). Tumors were verified histopathologically (9) and considered nonadvanced if they were 20 mm or less and advanced if greater than 20 mm in diameter (T2 to T4 in the TNM system) (2). Because the DCR did not record tumor size until 2004, we used DBCG data from 1980 to 2004. When we compared the DBCG database with the DCR after 2004, tumor size was not registered in 8% to 10% of the tumors in the DBCG database. We therefore used the DCR data after 2004. However, the DCR was missing the tumor size for 4% to 5% of tumors; thus, we excluded these tumors from our analyses. Registration of ductal carcinoma in situ (DCIS) was not mandatory in the DCR and DBCG before 2008 and 1989, respectively. Therefore, DCIS data were from the DBCG database for the entire observation period and are presented for the screened age group only because DCIS is mainly detected through screening. Statistics Denmark was the source of population size (10). Breast Cancer Screening in Denmark Organized breast cancer screening programs began in different regions at different times (Copenhagen in April 1991, Funen in 1993, and Frederiksberg in July 1994) and covered approximately 20% of the population (11). From late 2007, the remaining regions gradually introduced screening, but coverage was still incomplete in 2014 (12). Women aged 50 to 69 years were invited by mail to screening at a specified date and time. Screening was offered biennially and included 2-view mammography in the first round and 1-view mammography at subsequent rounds, except for women with dense breasts who always received 2-view mammography. The program did not include clinical breast examination. Screening results were mailed to women and their general practitioners within 10 working days. Women with abnormal results were referred to specialized units for additional testing. Participation rates in subsequent screening rounds were 62% in Copenhagen and 82% in Funen (11, 13). Danish women generally do not seek screening mammograms outside the organized program (14). Statistical Analysis We used a beforeafter approach to define screening and nonscreening areas. The screening areas were Copenhagen, Frederiksberg, and Funen, and the remaining 80% of Denmark was the nonscreening area. We merged Frederiksberg and Copenhagen because Frederiksberg is geographically surrounded by Copenhagen (43000 women in the first round collectively). For Copenhagen and Frederiksberg, the prescreening period was 1980 to 1990 and the screening period was 1991 to 2010. For Funen, the prescreening period was 1980 to 1993 and the screening period was 1994 to 2010. Similarly, for the areas without screening programs, the before period (prescreening period for the control group) was from 1980 to 1990 and the after period (screening period for the control group) was from 1991 to 2010, except for women aged 50 to 69 years in which the after period was from 1991 to 2007 (because of the beginning of national rollout). Association of Screening With Incidence of Advanced Cancer We used Poisson regression to analyze trends in incidence of advanced and nonadvanced tumors, adjusted for 5-year age groups, and stratified by screening and nonscreening areas. We compared incidence rates (IRs) of nonadvanced and advanced tumors in screening and nonscreening areas and calculated the annual percentage changes before and after screening. In the analysis of women aged 50 to 69 years in the nonscreening areas, we censored data for nonadvanced tumors in 2007 when national screening started. The IR, IR ratios (IRRs), and IR differences with 95% CIs were used to compare rates before and after screening for each age group (35 to 49, 50 to 69, and 70 to 84 years) in screening and nonscreening regions. We used Stata SE, version 14.0 (StataCorp). With Excel, version 14.0 (Microsoft), we produced graphs showing the 3-year moving average incidence of advanced and nonadvanced tumors and marked exact yearly IRs. Estimating Overdiagnosis In our first approach, we calculated the IR of advanced and nonadvanced tumors in the before and after periods for nonscreening and screening areas among women aged 50 to 84 years; the number of tumors before and after in the nonscreening and screening areas was calculated by multiplying the IR by the number of women aged 50 to 84 years living in Denmark in 2010. The number of overdiagnosed tumors was the difference between the number of tumors in the screening areas (afterbefore) and the nonscreening areas (afterbefore). We estimated overdiagnosis using the number of tumors among nonscreened women not being screened (nonscreening areas; standardized to the Danish population in 2010) in different age groups (50 to 69 and 50 to 84 years) as the denominators. We estimated overdiagnosis of invasive tumors only and invasive tumors and DCIS combined. This approach accounts for a reduction in the incidence of cancer due to earlier diagnosis in women no longer screened and for increasing incidence trends over time not related to screening. We included the prevalence peak in the screening areas because follow-up after the introduction of screening was longer than 10 years and estimates of the average lead time are 1 to 5 years (15). Thus, we allowed for sufficient follow-up after the first screening round to observe the expected decrease in incidence after an initial peak. (See the Supplement, for the formula and an example.) Supplement. Data Supplement In our second approach, we analyzed trends in advanced and nonadvanced cancer in the screening and nonscreening areas among women younger (35 to 49 years) and older (70 to 84 years) than those included in the program and compared these trends with those in women in the screening age range (50 to 69 years). This accounted for regional differences unrelated to screening. We found similar patterns in trends of advanced cancer among women eligible and ineligible for screening. The relative increase of advanced breast tumors was higher in the nonscreening areas than in the screening areas for women aged 35 to 49 years (IRR, 1.61 and 0.78), 50 to 69 years (IRR, 1.46 and 0.96), and 70 to 84 years (IRR, 1.81 and 1.25). We also found no compensatory decrease in the incidence of invasive cancer in women aged 70 to 84 years who were no longer offered screening. Thus, we concluded that screening was not associated with a reduction in the incidence of advanced cancer and used the incidence increase for nonadvanced tumors in women aged 50 to 69 years to calculate overdiagnosis. (See the Supplement for the formula and an example.) This study was exempt from institutional review board approval. Role of the Funding Source This study received no external funding. Results In 2010, the Danish population included 1420701 women aged 35 to 84 years. The DBCG database included 94932 women aged 35 to 84 years diagnosed with invasive breast cancer (n= 90665) or DCIS (n= 4267) from 1980 to 2010, whereas the DCR included data on 105994 women with invasive tumors and DCIS until 2011. The difference was mostly because of more tumors registered in the period before 1990. The Supplement Table shows the number of tumors (advanced and nonadvanced) and person-years in each age group for screening and nonscreening areas. Incidence of Advanced Cancer Figure 1 shows trends in the incidence of advanced cancer over time by age group. Table 1 shows the annual percentage of change before and after screening in screened and nonscreened areas by age group. Table 2 shows IRs, IR differences, and IRRs for screened and nonscreened areas by age group. Figure 1. Three-year moving averages of advanced breast tumors (>20 mm) in women aged 35 to 49 y (A), 50 to 69 y (B), and 70 to 84 y (C). The vertical dotted lines indicate the year of introduction of breast screening in Copenhagen (1991), Funen (1994), and the remaining regions in Denmark (2007). (See Supplement Figure 4, for separate incidence rates for Copenhagen and Funen.) Figure 1. Continued Table 1. Annual Percentage of Change in Incidence (95% CI) in the Screening an

Is birth history the key to highly educated women's higher breast cancer mortality? A follow-up study of 500,000 women aged 35-54.

A positive relationship has been found between high levels of education and breast cancer mortality. The aim of our study is to determine if the educational gradient in breast cancer mortality persists after adjustment for reproductive history. Register data including the total adult population in Norway were used. A total of 512,353 Norwegian women 35–54 years of age at the Norwegian Census in 1990 were followed with respect to breast cancer deaths until December 31, 2001. The analysis included 2,052 breast cancer deaths in 5.6 million person years. Educational differences in breast cancer mortality were analysed using Cox regression. The age adjusted relative risk of dying from breast cancer for women with >12 years of education compared to women with <10 years was 1.25 (95% confidence limits [CI] = 1.10–1.41). Adjustment for age at first birth with nulliparous as reference category reduced this difference to 1.08 (95% CI = 0.95–1.23). For parous women, age at first birth explained all the educational difference in breast cancer mortality. Among nulliparous women there was a larger positive educational gradient in breast cancer mortality than among parous women (relative risk [RR] = 1.57, 95% CI = 1.15–2.13), indicating that there were differences in other confounders than birth history among the childless.

Estimation of lead time and overdiagnosis in breast cancer screening

Sir, It is reported that 7 years after the start of screening in Sweden, the incidence of invasive breast cancer in the screened age group was 69% higher than expected (Duffy et al, 2008). After adjustment for a lead time of 2.4 years and for the increased use of hormone replacement therapy, they found 39% excess. We have reservations about adjustments for lead time (Zahl et al, 2008). Correction for lead time should only be done for those cancers that would have been diagnosed at a later time in the absence of screening. But many screen-detected cancers would not have come to the women’s attention in their remaining lifetime, if they had not attended screening and are by definition overdiagnosed. If these cancers are not excluded from the calculation of lead time, the lead-time distribution will be artificially rightskewed, and the average lead time will appear to be much longer than it really is. Duffy et al, (2008) did not exclude such cancers from the calculation of lead time, but counted all excess cancers detected in a randomised trial as advanced diagnoses, when some of them were in fact overdiagnosed cases. They have therefore overestimated the lead-time effect. Duffy et al (2008) recognise that the detection of non-overdiagnosed cancers should give rise to a drop in incidence when the women leave the screening programme. Quantifying such a compensatory drop is a less bias-prone method for adjusting the incidence increase for leadtime effects, as it makes no assumption about the average lead time. It is also very simple to use (see, e.g., Zahl et al, 2004). We have done a systematic review of incidence trends in countries with organised mammography screening that presented data on breast cancer incidence for both screened and nonscreened age groups for at least 7 years before screening and 7 years after screening had been fully implemented (Jorgensen and Gotzsche, 2008). We were able to include data from United Kingdom, Canada, Australia, Sweden and Norway. We found that compensatory drops in the older age groups were small or absent, although major drops would have been expected if the lead time of 2.4 years were true. When we adjusted for these drops, we found 36% overdiagnosis of invasive breast cancer, in good agreement with the results of Duffy et al (2008), and 51% when we included carcinoma in situ. Duffy et al (2008) mention that after prolonged follow-up of the Malmo randomised screening trial, an overdiagnosis of 7–8% was reported and they find this estimate considerably more plausible than their own estimate of 39%. However, they have overlooked that the former estimate is seriously flawed (Zahl et al, 2008). There was substantial opportunistic screening in the control group and after adjustment for this, the overdiagnosis estimate in the Malmo trial is 24% (Gotzsche and Jorgensen, 2006). It is asserted that after adjustments, overdiagnosis estimates will be smaller than many rates quoted in the past (Duffy et al, 2008). We disagree, as most ‘rates quoted in the past’ have been too small (Gotzsche and Nielsen, 2006). On the basis of randomised trials, the overdiagnosis is 30% (Gotzsche and Nielsen, 2006), and it becomes 44% if adjusted for opportunistic screening in the control groups (Jorgensen and Gotzsche, 2008). The low rates in the past have been too low for the very reason that they have been based on flawed lead-time models (Jorgensen and Gotzsche, 2008; Zahl et al, 2008). We believe the most reliable estimate for organised mammography screening is 51% (Jorgensen and Gotzsche, 2008). This means that one in three breast cancers detected in a population offered organised mammography screening are overdiagnosed. Many women are therefore harmed substantially by screening, as practically all detected carcinoma in situ cases and invasive cancers are treated, at great physical and psychological costs.

Chain of care for patients who have attempted suicide: a follow-up study from Bærum, Norway

BackgroundIndividuals who have attempted suicide are at increased risk of subsequent suicidal behavior. Since 1983, a community-based suicide prevention team has been operating in the municipality of Bærum, Norway. This study aimed to test the effectiveness of the teams interventions in preventing repeated suicide attempts and suicide deaths, as part of a chain of care model for all general hospital treated suicide attempters.MethodsData has been collected consecutively since 1984 and a follow-up was conducted on all individuals admitted to the general hospital after a suicide attempt. The risk of repeated suicide attempt and suicide were comparatively examined in subjects who received assistance from the suicide prevention team in addition to treatment as usual versus those who received treatment as usual only. Logistic regression and Cox regression were used to analyze the data.ResultsBetween January 1984 and December 2007, 1,616 subjects were registered as having attempted suicide; 197 of them (12%) made another attempt within 12 months. Compared to subjects who did not receive assistance from the suicide prevention team, individuals involved in the prevention program did not have a significantly different risk of repeated attempt within 6 months (adjusted OR = 1.08; 95% CI = 0.66-1.74), 12 months (adjusted OR = 0.86; 95% CI = 0.57-1.30), or 5 years (adjusted RR = 0.90; 95% CI = 0.67-1.22) after their first recorded attempt. There was also no difference in risk of suicide (adjusted RR = 0.85; 95% CI = 0.46-1.57). Previous suicide attempts, marital status, and employment status were significantly associated with a repeated suicide attempt within 6 and 12 months (p < 0.05). Alcohol misuse, employment status, and previous suicide attempts were significantly associated with a repeated attempt within 5 years (p < 0.05) while marital status became non-significant (p > 0.05). With each year of age, the risk of suicide increased by 3% (p < 0.05).ConclusionsThe present study did not find any differences in the risk of fatal and non-fatal suicidal behavior between subjects who received treatment as usual combined with community assistance versus subjects who received only treatment as usual. However, assistance from the community team was mainly offered to attempters who were not receiving sufficient support from treatment as usual and was accepted by 50-60% of those deemed eligible. Thus, obtaining similar outcomes for individuals, all of whom were clinically judged to have different needs, could in itself be considered a desirable result.