Letters
Risk of connective tissue disease among women with breast implants
Study adds nothing to knowledge of processes of tissue injury
induced by silicone Authors should have made better use of matched control groupAuthors' replyStudy adds nothing to knowledge of processes of tissue injury induced by silicone
EDITOR Table 3 is confusing because the observed numbers of cases are compared
with the expected numbers, derived from the standardised hospitalisation ratios. The data can mean only that in Sweden the rate
of admission to hospital for rheumatic disease is the same whether a
patient has silicone implants or not. Few patients go to hospital for
rheumatic assessment in the United States. We have admitted none for
rheumatic disease over the past two years from our silicone clinic; not
all removals of implants are done in hospital.
Adjustments for pre-existing diagnosis or miscoding do not seem to fit
between tables 3 and 5; the total of individual defined disorders
exceeds that for all patients by about a third, which is the frequency
of overlap syndrome. Overlap syndrome has been ignored in this study
despite its importance in the classification of rheumatic disorders.
Data adjustments are necessary to ensure the accuracy of the tables,
but they may not show the biological meaning of the data. For example,
average follow up does not indicate how skewed the clinical assessment
point is with respect to the average. Brawer has shown a time dependent
crossover at 5-6 years between change in clinical status and cumulative
rupture rate in 300 patients with siliconosis.3
The study is short when one considers the immunopathic reaction cycle
before autoimmune conversion; the T cell reaction peaks 10-12 years
after implantation and then declines, reaching effective quiescence
more than 20 years after implantation unless immunomodulatory treatment
or removal of the implants intervenes.5
The report adds nothing to discerning the processes of tissue injury
induced by silicone, is too short to detect autoimmune conversion,
ignores the atypical nature of siliconosis seen clinically and in
laboratory tests, and is close to the logical fallacy of taking an
absence of evidence as evidence of absence.
a
Professor Shanklin has acted on behalf of a patient in
one case recently; Professor Smalley has not acted in any relevant cases since mid-1995.
Authors should have made better use of matched control group
EDITOR Their finding of a relative risk of 0.8 for being admitted with a
connective tissue disorder is derived from the whole breast implant
group (cosmetic and reconstruction) compared with the breast reduction
group. Thus, having selected a control group, they fail to use it
appropriately or to provide data allowing others to do so. From the
data they do present, one can compare the cosmetic implant group as a
whole with the breast reduction group, although this is imperfect
because some patients from the cosmetic implant group are unmatched.
This comparison gives an unadjusted relative risk of 1.2 for patients
with breast implants later developing connective tissue disorders
requiring admission to hospital. Although the 95% confidence interval
includes 1.0, this may be a less palatable figure to the funders than
the relative risk of 0.8 misleadingly quoted.
I urge the authors to state the relative risk and confidence intervals
(from their matched patients) of admission to hospital with connective
tissue disorder after receiving cosmetic breast implants compared with
breast reduction. They should also present data on the subgroup with
silicone gel implants, using the matched controls for that subgroup.
The authors should emphasise the size of relative risk of developing
connective tissue disorders which they can exclude with 95%
confidence. They could allay public concern (if appropriate) by quoting
mean or maximum absolute risks and comparing these with other absolute
risks from the literature Authors' reply
EDITOR Shanklin and Smalley suggest that silicone would not be expected to
induce conditions serious enough to require admission, but many of the
original hypotheses about breast implants involved debilitating
illnesses. The authors claim that our study was "five years old at
publication." This is untrue; we only recently completed the
fieldwork. Their confusion about tables 3 and 5 may stem from a failure
to understand the concept of a standardised hospitalisation ratio or
their lumping together fibromyalgia with definite connective tissue
disease. We considered two broad diagnostic entities Atherton suggests that our analysis was biased by the source of
funding. We prefer to have the study judged on its methods rather than
on non-scientific innuendo. The cohort study approach that we used is
common epidemiological practice and offered the advantage of a
standardised ascertainment and follow up of patients. Atherton objects
to our use of the entire implant cohort instead of the cosmetic implant
subcohort in making the direct comparison with the breast reduction
cohort. Although in the design stage the patients who had breast
reduction were matched to the patients with cosmetic implants, in the
analysis stage While subject to the limitations we ourselves highlighted, our
study provides systematic information obtained in an area of the world
much less affected than others by publicity about adverse effects of
implants. It shows that women with implants experience rates of serious
connective tissue disease similar to those of other women in Sweden.
In a study that was five years old at publication Nyrén et al
claim to show "no evidence of association between breast implants and
connective tissue disease," using classical rheumatic diseases as end
points.1 This report is less comprehensive than that by
Gabriel et al, although comparable in scope and shortcomings of
definitions.2 It lacks data on rupture (the prevalence
rises with time3); rupture enhances the reaction to gel
filled devices.4 Gabriel et al estimated the minimum
population necessary for risk assessment to be 62 000 subjects with
implants and 124 000 controls.2
David L Smalley
Department of Pathology, University of Tennessee, Memphis, TN
38163, USA
In view of their acknowledged funding from Dow-Corning
Corporation, Nyrén et al should have been more careful to avoid even
subtle bias in the presentation of their results.1 They analysed two groups of patients with breast implants (cosmetic and
breast reconstruction groups) and a carefully matched control group for
the patients with cosmetic implants, consisting of women who had breast
reduction surgery. The most important comparison should have been
between the cosmetic breast implant group and the matched control
group;making this comparison is the reason for selecting that control
group.
for example, the risk of developing lung
cancer if one smokes.
Institute of Infections and Immunity, University Hospital,
Nottingham NG7 2UH
We studied over 7000 women, and although this was one of the
largest cohorts of women with implants and the follow up was one of the
longest, we clearly stated that the outcomes evaluated (admissions to
hospital) represented only the more serious illnesses from definite
connective tissue disease. No evidence of increased risk of admission
for these conditions was associated with implants.
definite connective tissue disease, and related conditions including
fibromyalgia. Shanklin and Smalley infer that considerable overlap
syndrome occurred; in fact the prevalence of multiple conditions was
small and similar among those with breast implants and those who had breast reduction surgery.
by matching for several variableswe could include all
patients with implants. Nevertheless, if the patients with cosmetic
implants were compared with those who had breast reduction the
resulting relative risk for all definite connective tissue disease
would be 1.0 (95% confidence interval 0.6 to 1.8). Because we had
information on the silicone content of only a sample of the implants,
we did not calculate risk ratios according to implant type.
Hans-Olov Adami
Li Yin
Staffan Josefsson
Department of Medical Epidemiology, Karolinska Institutet, PO
Box 281, S-171 77 Stockholm, Sweden
Joseph K McLaughlin
William J Blot
John D Boice Jr
International Epidemiology Institute, Rockville, Maryland, USA
Martin Engqvist
National Board of Health and Welfare, Stockholm, Sweden
Lars Hakelius
Department of Plastic and Hand Surgery, University Hospital,
Uppsala, Sweden
© BMJ 1998
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Relevant Article
- Risk of connective tissue disease and related disorders among women with breast implants: a nation-wide retrospective cohort study in Sweden
- Olof Nyrén, Li Yin, Staffan Josefsson, Joseph K McLaughlin, William J Blot, Martin Engqvist, Lars Hakelius, John D Boice Jr, and Hans-Olov Adami
BMJ 1998 316: 417-422.[Abstract] [Full Text] [PDF]
This Article
-
Extract
- Respond to this article
- Alert me when this article is cited
- Alert me when responses are posted
- Alert me when a correction is posted
Services
- Email this article to a friend
- Find similar articles in BMJ
- Find similar articles in PubMed
- Add article to my folders
- Download to citation manager
- Request Permissions
Citing Articles
Google Scholar
PubMed
Related Content
Bookmark with
What's new
Latest blogs and video
Services
Tools
Online poll
Resources
Rapid responses for this article
There are no rapid responses for this article.Most read
Print issues
