RESULTS OF RANDOMIZED TRIALS FOR WOMEN AGE 40 - 49 by rt3463df

VIEWS: 5 PAGES: 19

									Does Mammographic Screening reduce
  the mortality from breast cancer?


     Andy Coldman, Yulia D’yachkova,
         Norm Phillips, Lisa Kan,
              Linda Warren
Q: What is screening Mammography?


A: Screening is the use of mammography to
   identify women likely to have breast
   cancer who do not have signs or symptoms
   of the disease.
Q: How do we tell if screening works?


A: Randomized control trials (RCT). Women
   without cancer are randomized to receive
   (or not) several “rounds” of screening and
   then followed and their mortality from
   breast cancer compared.
Q: So if RCT’s will give us the answer
   haven’t there been any used to test
   screening mammography?


A:   Yes there have, but they don’t all give
     exactly the same answer.
Results of Randomized Control Trials in
           Women Age 50 - 74
      STUDY      MORTALITY           CONFIDENCE
                 REDUCTION              INTERVAL
EDINBURGH           20%             - 17% to + 46%
MALMO               21%             - 24% to + 49%
KOPPARBERG          39%               11% to 58%
OSTERGOTLAND        31%                 6% to 50%
NBSS2                3%             - 52% to 38%
HIP                 35%                 8% to 54%
STOCKHOLM           35%               - 8% to 35%
GOTHENBURG           9%             - 55% to 47%
META ANALYSIS       23%               13% to 31%
SUMMARY
                    ** Kerlikowshe et al: Jama 273: 149-154
Results of Randomized Trials for Women
               Age 40 - 49
      STUDY      MORTALITY            CONFIDENCE
                 REDUCTION             INTERVAL
EDINBURGH           19%                 - 67% to 46%
MALMO               36%                 - 11 % to 55%
OSTERGOTLAND        - 2%                 - 77% to 41%
KOPPARBERG          33%                   5% to 63%

NBSS1              - 14%                - 56% to 17%
HIP                  23%                - 11% to 47%
STOCKHOLM            - 1%              - 101% to 49%
GOTHENBURG           44%                   2% to 68%
META ANALYSIS        18%                   5% to 29%
SUMMARY

                            ** Herdrick et al: JNCI 22:87-92
Q: Haven’t the results of these trials
   been challenged and disproven?
A: A small number of authors have been critical of
   these studies and this has received a lot of media
   attention. The major criticisms have been:
        1. The multiple reports from individual trials
           provide inconsistent numbers.
        2. That randomization was flawed in several
           of the trials.
        3. That the trials used deaths from breast
           cancer, rather than deaths from all causes
           in measuring outcome.
1. The multiple reports from individual
   trials provide inconsistent numbers.

   • It is true that numbers from each report
     from a trial does not have exactly the
     same numbers. This has been generally
     been caused by blinded review processes
     which have changed the eligibility of
     some cases. In all trials the magnitude of
     such changes is small so that results are
     unaffected.
2. That randomization was flawed in
   several of the trials
  • Not all trials used individual
    randomization but randomized districts
    and then included all eligible women in
    those districts into the trial.
    Consequently the two arms (screening v
    non-screening) were not exactly
    balanced. Whilst true the differences
    were extremely small and were more
    often in the direction not favouring
    screening.
3. That the trials used deaths from
   breast cancer, rather than deaths
   from all causes in measuring outcome.
  • Although breast cancer is an important
    cause of death most women in screening
    studies die of other causes. Thus the
    effect of screening on breast cancer gets
    lost when all causes are used.
    Nevertheless a recent analysis of some
    trials show that all-cause mortality is
    reduced.
• These criticisms have been reviewed
  by a number of expert committees
  and have been concluded to be
  unfounded.

• Every western country recommends
  mammography screening in women
  only the age ranges differ.
Q: What is the situation in BC
regarding mammographic screening?
A: •Screening began in 1988.
   •Screening is offered across the province in
       40 affiliated centres.
   •Screening offered to women aged 40-79 and
       targeted to women 50-74.
   • 50% of women 50-74 are screened every 2
       years.
   •230,000 screens will be performed in 2003.
Trends in Breast Cancer Incidence and Mortality
           by Age Between 1974 - 98
                                                                                                       All Ages with 95% CI
                                                                                                                    With  10
                              Breast (Female)                                                          Age 0-44
                                                                                                       Age 45-64    cases per
                                                                                                       Age 65+      year
                                     Females
                              200                                            200                                           200
Relative rate (% log scale)




                                      Incidence                                       Mortality

                              150                                            150                                           150

                              125                                            125                                           125

                              100                                            100                                           100

                               80                                            80                                            80
                               70                                            70                                            70

                               60                                            60                                            60

                               50                                            50                                            50

                                    1974   1980       1986   1992     1998         1974    1980      1986   1992    1998


                                                  Year of diagnosis                               Year of death
Numbers of Cancers Detected at
     Screening by year
         1993 - 2002
    1993          377
    1994          478
    1995          522
    1996          607
    1997          734
    1998          860
    1999          887
    2000          852
    2001          909
    2002         1,019
Stage Distribution of Cancers in Screened
   and Un-screened British Columbia
             Women in 1995

   Stage      Screened       Unscreened
            #         %     #         %
   I       543        62   447        39
   II      259        30   529        46
   III     69          8    91        8
   IV       3        0.3    77        7
   Total   874       100   1144      100
Observed Percentage Point Increase in 5 - Year
Relative Survival by Stage for Cases Referred to
  BCCA 1987 - 96 and All Cases in Province

   Stage                       % Increase in
                                 Survival
   I                                4
   II                               5
   III                              5
   IV                               0
   All Provincial Cases             7
   Predicted if No Screening        4
       5 Year Relative Survival
Unscreened versus Screen Detected
   Cancers Referred to BCCA
           1990 - 1996
 Stage      Unscreened    Screen
                         Detected
 I             97%        100%
 II            83%         95%
 III           57%         69%
 IV            21%         32%
Observed and Expected Breast Cancer mortality rates
by time since first screen for women screened at least
              once between ages 50-74

             Observed
             Expected




         0      2       4      6      8      10
Conclusions

  •Yes mammographic screening has been
  demonstrated to reduce mortality in RCT’s.
  •Screening is detecting increasing numbers
  of cancers in BC.
  •Screen detected cancers have a better
  prognosis.
  •Breast cancer mortality is lower in BC
  women who use screening.

								
To top