silicon dioxide stimulation for women with rheumatic dis-
orders was indistinguishable from the normal controls
{P = 0.3577). The average stimulation index for silicon diox-
ide in the rheumatic disorder patients was 7.1 (Fig. 1).
Among the 220 normal adults tested in the expanded
study, the mean stimulation index (SI) was 10.0 (Fig. 2). To
assess the distribution of SI for the implant patients, we used
2.5 times the mean of normal controls, 5.0 times the mean.
TABLE 2 , Mean counts/min of implant groups separated by level of response
compared to normal controls (expanded study)
Unstimulated
Con A
Silicon dioxide
Group
CPM
CPM
CPM
Normal controls
31
8690
281
(n - 220)
SI, 0-25
33
7897
500
{n = 82)
SI, 25-50
34
8740
1308
(n - 171)
SI, 50-100
28
8357
2101
(n = 316)
SI, >100
24
9264
3912
(« - 373)
IMMUNE RESPONSE IN SILICONE DISEASE
425
121
RESEARCH COMMUNICATION'
f
â–
H ASVMPTOMATIC
«
^ 4
^n
^3^—,
1 i
U— LJ
JU
lOJO 20-2S 25-50 50-100
Figure 3. Stimulation indexes of 34 asymptomatic silicone breast
implant patients after lymphocyte stimulation with silicon dioxide
(silica, expanded phase).
and 10.0 times the mean as points of significant levels. Nine
hundred forty-two implant patients were tested; 860 (91.3%)
had SI above 25 (2.5 times the control mean), which we es-
tablished as the threshold for positivity. One hundred
seventy-one (18.2%) of the symptomatic patients had SI be-
tween 25 and 50; 316 (33.5%) has SI from 50 to 100; and 373
(39.6%) had SI greater than 100, more than 10 times the
control mean (Fig. 2). As shown in Table 2, the mean un-
stimulated counts per minute for all groups had little vari-
ance as did the mean for Con A counts per minute. The raw
cr- f,*'.
"^^
Figure 5
giant cells m mammary prosthesis tapsular tissue 1 lie .suljstagc has
been set down slightly to enhance the optical quality of the silicone,
which does not stain (arrows). Hematoxylin and eosin. Original
magnification, 100 x.
^^
Figure 4. Intense lymphocytic infiUration about vaned globular
pockets of silicone, mammary prosthetic capsular tissue Hematox-
ylin and eosin. Original magnification. lOOX.
counts for implant patients after stimulation with silicon di-
oxide varied from 500 to 3912 as reflected in the range of SI.
The SI distribution for asymptomatic women is shown in
Fig. 3. Twenty-two (64.7%) of these women were above the
index of 25 and 19 of the 22 (86.4%) were below an index
of 100. As shown in Fig. 4, the lymphocyte reaction is occa-
sionally very severe in human implant capsules with globular
silicone throughout the tissue; sometimes the reaction is es-
sentially pure foreign body granuloma (Fig. 5). Both reac-
tions to silicon dioxide are commonplace in human and
animal material; studies with fumed amorphous silica have
shown it to cause a more severe lymphocyte reaction (15).
DISCUSSION
The data demonstrates clearly that women with silicone
mammary implants develop a cell-mediated immunopathic
response to silica. Previous observations confirmed the
specific cellular response was CD3* T lymphocytes (16). The
lack of correlation between the T cell SI and the length of
prosthesis exposure time is suggestive of variable reactivity
among recipients over time and supports the belief that sen-
sitization occurs early. In assessing patients with leakage or
ruptures confirmed by ultrasound, magnetic resonance im-
aging, or surgical observation, we found the amount of leak-
age did not correlate with the level of response. Among 64
patients in the 0-25 range of SI, 33 (51.6%) had confirmed
leaks. In the remaining groups with confirmed leaks, 60 of
148 (40.5%) patients had SI between 25 and 50; 126 of 280
(45.0%) patients were in the range 50-100; and 122 of 303
426 Vol. 9 March 1995
The FASEB lournal
SMALLEY ET AL.
122
TABLE 3. Principal ingrtdimts of representative envelope ("skeW) maUrial
in use. 1967-1992
Envelope maierial
bv siotk number'
MDF-077
Q7-2423
Dimethyl methylvinyl siloxane,
dimelhylvinyl-tcrminaled
Dimelhyl siloxane, dimelhylvinyl-
icrminatcd
â– "Amorphous silica"
Mcthylhydrogcn siloxane
Dimcthylhydrolyaic
Hcxamethyldisiloxane
Hexamclhyldisilazane
Dimethyl methylvinyl siloxane
Dimethyl methylhydrogen siloxane
Chloruplatinic acid
Methylvinyl cyclosiloxane
63.29%
6.32
0.00
57.52
26.50
26.18
0.00
4.77
0.00
3.08
0.00
1.90
8.83
0.08
0.00
0.08
1.07
0.00
0.15
0.07
0.15
0.00
'Dow Corning Corporation.
(40.3%) patients had SI greater than 100. These findings
suggest exposure to low molecular weight gel bleed and the
outer silicone shell (Table 3). Mitogenic studies among the
implant patients confirm that normal stimulation occurs and
that it is not different from the normal controls. None of the
initial patients had SI below 25; however, in the expanded
study nearly 10% of symptomatic patients showed no in-
creased response to silicon dioxide This may represent a
group of women postexplant with immune quiescence or
they may be nonresponders to silicon dioxide with the possi-
bility their symptoms are due to some other cause or causes.
A large percentage (91.3%) of symptomatic implant pa-
tients did demonstrate T lymphocyte response to silicon di-
oxide. Silica accounts for a quarter of the envelope (Table 3)
(17), and in vivo degradation of silicone has now been estab-
lished by nuclear magnetic resonance (18). In addition, clear
evidence of spread of silicone and its metabolites to nearly
every major organ may facilitate the immune reaction and
account for many specific symptoms among implant patients (19).
The demonstration of specific T lymphocyte response to
silica is in agreement with the comparative studies recently
reported by Ojo-Amaize et al. (20). These authors studied
lymphocyte responses to elemental silicon, unspecified sili-
cone, silica, and various other metals, including beryllium,
chromium, and nickel. Silica caused a consistently more
vigorous result than either silicon or silicone as well as all
other substances that were compared (20).
Asymptomatic implant patients clearly show a different
distribution of SI. Of the 34 tested so far, 22 had SI below
100. This suggests some patients may be immunologically
reactive at low levels and have not begun to manifest sym-
ptoms. Another possibility is the data are showing a group
of nonresponders with the need for assessment for genetic
markers for tolerance.
The present study confirmed that lymphocytes from
women exposed to silicone gel mammary implants can be
antigenically stimulated by silicon dioxide. Accordingly, hu-
man tissue reactions to substances in or from the implants
follow the expected immunopathic sequence of processing by
RESEARCH COMMUNICATION
macrophages, .sometimes leading to granuloma formation,
and presentation to lymphoid centers for specific T cell
f*roduction. This study shows a T lymphocyte response to
silicon dioxide (silica), which likely contributes to tissues
changes seen pathologically and to the spectrum of clinical
silicone-associated disease. [fTl
REFERENCES
1. Shons. A, R.. and Schubert. \V (1992) Silicone- breast implants and im-
mune disea.ie. Ann. Blast. Surg 28, 491-499
2 Varga. J., Schumacher, R., and Jimenez, S. A- (1989) Systemic sclerosis
after augmeniaiion mammoplastv with silicone implants. Ann /nt Med
111, 377-383
3. Spiera, H.. and Kerr, L, D. (1993) Sclerodrrrr.d following silicone im-
plantation: a cumulative experience of 11 cases. y. Rheumatol 20, 938-961
4. Silver. R M., Sahn. E. E.. Allen. J A. Sahn. S.. Greene. W. Maize.
J, C, and Garen, P. D (1993) Demonstration of silicon in sites of con-
nective tissue disease in patients with silicone-gel breast implants. Arch
Derynatol 129. 63-68
5. Rudolph. R . Abraham. J-. Vecchione. T. Guber. S,. and Woodward.
M. (1978) Myofibroblasts and free silicon around breast implants. Plasl
ReconstTuct Surg. 62, 223-229
6. Baker. J L.. LeVier. R R., and Spielvogel. D. E. (1982) Pbsiiive
identification of silicone in human mammary capsular tissue, f^l.
Reconstruct Surg 69, 56-60
7 Thomsen. J L . Christensen. L.. Nielsen. M.. Brandt. B., Breitling.
V, B.. Felby. S,. and Nielsen, E. (1990) Histologic changes and silicone
concentrations in human breast tissue surrounding silicone breast
prostheses. PlasI Reconstruct Surg 85, 38-41
8, Shanklin. D. R (1991) l^te tissue reactions to silicone and silica. In 5t/i-
fon^m A/«/Ka/D«'ifa(Stratme\'er. M. E. ed) pp. 103-125. USHHS/FDA.
Conference Proceedings. Baltimore
9. Shanklin, D, R, (1993) Silicone-associated diseases: tissue findings. South
Med J 86. S104
10 Goldblum. R M . Pelley. R, P.. O'Donell. A A . Pyron. D. and Heg-
gers, J. P (1992) Antibodies to silicone elastomers and reactions to vcn-
triculoperitoneal shunts. Lancet 340, 510-513
11 Wolf. L E . Lappe. M , Peterson. R D. and Ezrailson. E G. (1993)
Human immune response to polydimethylsiloxane (silicone): screening
studies in a breast implant population. FASEB J 7, 1265-1268
12, Bridges, A. J.. Conley. C. Wang. G,. Burns, D. E.. and Vasey. F, B,
(1993) A clinical and immunologic evaluation of women with silicone
breast implants and symptoms of rheumatic disease. Ann. Int. Med 118,
929-936
13, Geha. R. S.. and Merler. E. (1974) Response of human ihymus-derived
(T) and non-thymus-derived (B) lymphocytes to mitogenic stimulation
in vitro. Eur J Immunol 4, 193-199
14 Chilson. O P. and Kelly-Chilson. A. E. (1989) Mitogenic lectins bind
to the antigen receptor on human lymphocytes. Eur / Immunol 19,
389-396
15 Picha. G. J., and Goldstein, J. A. (1991) Analysis of the soft-tissue
response to components used in the manufacture of breast implants: rat
animal model. Plast. Reconstruct Surg Z7, 490-500
16 Smalley. D, L,. Shanklin. D. R.. Hall. M F. and Stevens. M V. (199.^)
Detection of lymphocyte stimulation by silicon dioxide, ^ Occupat Med
Ibxicol In press
17. D()w Coming Corporation (1968. 1979) Mammary implant material
formulations, MDF-077 and Q7-2423. in Silicone Breast Implant Team
Leaders' Report. Surgery Device Panel, U.S. Food and Drug Adminis-
tration, released 1993. pp 42. 51
IB Garrido, L , Pfleiderer. B.. Papiso\. M . and Ackerman. J. I. (1993) In
vivo degradation of silicones. Magn Reson Med 29, 839-843
19 Schmiterlow, C G,. and Sjogren. C (1975) The distribution of '*C-
labelled KABI 1774, Ada Pharmacol lixicol 36, 131-138
20 Ojo-Amaize, E A., Come. V, Lin. H-C . Brucker. R F. Agopian.
M. S.. and Peter. J B, (1994) Silicone-specihr blo<xJ lymphocyte
response in women with silicone breast implants. Clin Diag Lab Im-
munol I, 689-695
Received for publttatton August 22. 1994
Accepted Jor publication Sotrmber 2H. 1994
IMMUNE RESPONSE IN SILICONE DISEASE
123
Mr. Shays. Thank you, sir.
This is an interesting panel.
Dr. Gabriel.
Dr. Gabriel. Thank vou.
Mr. Chairman, memoers of the committee, my name is Sherine
Emily Gabriel. I'm a rheumatologist and associate professor of
medicine smd epidemiology at Mayo Medical School and the prin-
cipal investigator of the study entitled "Risk of Connective Tissue
Diseases and Other Disorders After Breast Implantation," which
was published in the New England Journal of Medicine on June 16
of last year.
Connective tissue diseases are autoimmune disorders such as
rheumatoid arthritis and lupus, which are characterized by inflam-
mation of the joints, skin, and internal organs, as you have heard.
Since 1962, approximately 900,000 North American women have
received silicone breast implants. An increased risk of connective
tissue diseases related to implants has been postulated in the med-
ical literature. This is an important concern for many of these
women.
Mr. Chairman, strong scientific evidence now indicates that there
is no link between breast implants and an increased risk of these
conditions. My testimony will focus on that evidence.
In order to establish whether breast implants cause connective
tissue diseases, it is not enough to note that some women with im-
plants have developed these conditions. Instead, it is necessary to
determine whether women with implants are developing these con-
ditions at a higher rate than women of the same age and health
who do not have implants.
The only way to determine whether the rate of these diseases is
higher among women with implants compared to women without is
to perform vmat is known as a controlled study. In the absence of
a control gfroup, that is, a comparison group of women without im-
plants, the conclusion that medical conditions among women with
implants are, in fact, caused by these devices is not scientifically
valid.
Until recently, the published medical literature describing the re-
lationship between breast implants and connective tissue disorders
consisted virtually entirely of case reports and case series; that is,
descriptions of one or more women with implants who subsequently
developed a variety of medical problems. Such reports contribute
virtually no scientifically valid information bearing on the question
of whether implants cause connective tissue diseases.
As you have repeatedly heard today, numerous controlled, sci-
entifically valid studies have now been published which specifically
address this question. As an aside. Dr. Brown earlier testified to
the relative strengths and weaknesses of case control and cohort
studies, citing a single case control study.
There are, in fact, several others which are referenced in my tes-
timony, including a multicenter case control study among 869
women with systemic sclerosis, recruited from three university-af-
filiated rheumatology clinics and 2,061 matched community con-
trols. The relative risk of that study was also not statistically sig-
nificantly different from unity.
124
Our study included virtually all Olmsted County women who re-
ceived breast implants between January 1, 1964, and December 31,
1991, a total of 749 women with implants who were compared to
1,498 community women who did not have implants. We found no
connection between breast implants and connective tissue diseases.
These results have been confirmed in numerous additional con-
trolled studies, as I have mentioned.
Most recently, a controlled study from Harvard also found no as-
sociation between silicone breast implants and an increased risk of
either connective tissue diseases or a list of 42 related symptoms
amone 121,700 registered American nurses followed since 1976.
And Uiese results also, as you have heard, were published in this
year's New England Journal of Medicine, this June.
To summarize, not one of these controlled epidemiologic studies
identified a link between breast implants and connective tissue dis-
eases. These remarkably consistent results prompted the govern-
ments of some countries to review this topic. For example, in De-
cember 1994, the medical devices agency of the Department of
Health in the United Kingdom reported their evaluation of the evi-
dence for an association between breast implants and connective
tissue diseases.
The resulting 60-page document, which included critical apprais-
als performed by an independent scientific expert advisory group,
concluded that "There is no evidence of an increased risk of connec-
tive tissue diseases in patients who have had silicone gel breast im-
plants and therefore no scientific case for changing practice or pol-
icy in the United Kingdom with respect to breast implantation."
In light of this overwhelmingly consistent accumulation of sci-
entific research, I respectfully recommend that the Food and Drug
Administration assemble an independent panel of scientific experts
who have no ties with industry and have not been involved in
breast implant litigation. This panel would carefully review the
available evidence specifically regarding an excess risk of connec-
tive tissue diseases among women witn breast implants and pro-
vide a public policy statement. It is my hope that this statement
will reduce some of the anxiety that women with implants feel re-
garding the future.
Finally, Mr. Chairman, this was not part of my prepared com-
ments, but there have been some comments made this morning
about research funding, so I would just like to clear the air on this
issue with respect to my own study. For my study, there were three
sources of funding.
Mayo Foundation provided the initial funding. We then success-
fully competed in a peer-reviewed research grant competition from
the Educational Foundation of the American Society for Plastic and
Reconstructive Surgeons. And, finally, the National Institutes of
Health provided the funding for the Rochester epidemiology project,
which is the underlying data resource upon which this study was
based.
I wanted to emphasize that my interest in the breast implant
issue has always been strictly scientific. And while we did receive
the grant from the Plastic Surgery Educational Foundation, the
study was independently conceived, designed, and implemented by
the research team under my direction.
125
In fact, the design was complete and the study was already
under way, using Mayo funds, before the grant was awarded.
The Plastic Surgery Educational Foundation had no input into
the design, analysis, or interpretation of the results, and they
placed no limitations or restrictions on the publication of the re-
sults. The bottom line, Mr. Chairman, is that we would have done
exactly the same study, exactly the same way, regardless of who
^nded it.
Thank you.
[The prepared statement of Dr. Gabriel follows:]
Prepared Statement of Sherine E. Gabriel, M.D., Mayo Cunic
Mr. Chairman, members of the committee, my name is Sherine Emily Gabriel. I
am a rheumatologist, and an Associate Professor of Medicine and Epidemiology at
Mayo Medical School. I am also the principal investigator of the study entitled ^isk
of Connective Tissue Diseases ana Other Disorders After Breast Implantation"
which was published the New England Journal of Medicine on June 16, 1994.^ Con-
nective-tissue diseases are autoimmune disorders such as riieumatoid arthritis and
lupus, which are characterized by inflammation of the joints, skin, and internal or-
gans.
Since 1962, approximately 1 to 2.2 million North American women have received
silicone breast unplants.^*^ An increased risk of connective-tissue diseases, related
to implants, has been postulated in the medical literature.*"^'* This is an important
concern for many of these women. Mr. Chairman, strong scientific evidence now in-
dicates that there is no link between breast implants and an increased risk of these
conditions. My testimony will focus on this evidence.
In order to establish whether breast implants cause connective-tissue diseases, it
is not enough to note that some women with implants have developed these condi-
tions. Instead, it is necessary to determine whether women with implants are devel-
oping these conditions at a higher rate than women of the same age and health who
do not have implants. The only wav to determine whether the rate of these diseases
is higher among women with implants compared to women without implants is to
perform what is known as a controlled study. In the absence of a control group, i.e.,
iGabriel SE, CFallon WM, Kurland LT, Woods JE, Beard CM, Melton LJ, III: Risk of connec-
tive-issue diseases and other disordere afler breast implantation. N Engl J Med 330(24): 1697-
1702, 1994.
'Kesaler DA: The basis of the FDA's decision on breast implants. N Engl J Med 326:1713-
1715, 1992.
^Independent Advisory Committee on Silicon-Gel- Filled Breast Implants: Summary of the re-
port on silicon-gel-filled breast implants Can Med Assoc J 147:1141-1146, 1992.
*Hito8hi S, Ito Y, Takehara K, Fujiba T, Ogata E: A case of malignant hypertension and
scleroderma after cosmetic surgery. Jpn J Med 30(1):97-100, 1991.
'Gutierrez FJ, Espinoza LR: Progressive systemic sclerosis complicated by severe hyper-
tension; Reversal aOer silicone implant removal. Am J Med 89(3):390-392, 1990.
^Varga J, Schumacher HR, Jimenez SA: Systemic sclerosis afler ugmentation manunoplasty
with silicone implante. Ann Intern Med lll(5>.377-383, 1989.
''Sahn EE, Garen PD, Silver RM, Maize JC: Scleroderma following augmentation
mammoplasty. Report of a case and review of the literature. Arch Dermatol 126(9): 1198-1202,
1990.
*Spiera H: Scleroderma afler silicone augmentation mammoplasty. JAMA 260(2):236-238,
1988
"Brozene SJ, Penske NA, Cmse CW. Espinoza CG, Vasey FB, Germain BF, Fspinoza LR:
Human adjuvant disease following augmentation mammoplasty. Arch Dermatol 124(9):1383-
1386, 1988.
^"Okano Y, Nishikai M, Sato A: Scleroderma, primary biliary cirrhosis, and Sjogren's syn-
drome afler cosmetic breast augmentation with silicone injection: A case report of possiole
human adjuvant disease. Am Rheum Dis 43(3):520-522, 1984.
^^Kumagai Y, Shiokawa Y, Medsger TA, Jr, Rodnan GP: Clinical spectrum of connective tis-
sue disease after cosmetic surgery. Observations on eighteen patients and a review of the Japa-
nese literature. Arthritis Rheum 27(1): 1-12, 1984.
^Kumagai Y, Abe C, Sldokawa Y: Scleroderma afler cosmetic surgery. Four cases of human
a<«uvant disease. Arthritis Rheum 22(5):532-537, 1979.
"van Nunen SA, Gabenby PA, Basten A: Post- mammoplasty connective tissue disease. Ar-
thritis Rheum 25(6):694-697, 1982.
^* Baldwin CM Jr., Kaplan EN: Silicon induced human adjuvant disease? Ann Plast Surg
10(4):270-273, 1983.
"Byron MA, Venning VA, Mowat AG: Post-mammoplasty human adjuvant disease. Br J
Rheumatol 23(3):227-229, 1984.
126
a comparison group of women without implants, the conclusion that medical
condidons among women with implants are, in fact, caused by these devices is not
scientifically valid. ^® Until recently, the published medical literature describing the
relationship between breast implants and connective-tissue disorders consisted vir-
tually entirely of case reports and case series, i.e., descriptions of one or more
women with implants who subsequently developed a variety of medical problems.
Such reports contribute virtually no scientifically valid information bearing on the
question of whether implants cause connective-tissue diseases.
Over the past 18 months, 7 controlled, scientifically-valid studies have been pub-
lished whicn specifically address this question. Our study included virtually all
Olmsted County, Minnesota women who received breast implants between January
1, 1964 and December 31, 1991; a total of 749 women with implants and 1498
women who did not have implants.^ We found no connection between breast im-
plants and connective-tissue diseases. These results have been confirmed in 6 addi-
tional controlled studies. ^'"^ Most recently, a controlled study from Harvard found
no association between silicone breast implants and an increased risk of either con-
nective-tissue diseases or related symptoms among 121,700 registered American
nurses followed since 1976. ^'^ These results were published in the New England
Journal of Medicine in June of this year. To summarize, not one of these 7 con-
trolled epidemiological studies identified a link between breast implants and connec-
tive-tissue diseases.
These remarkably consistent results prompted the governments of several Euro-
pean countries to issue policy statements on this topic. For example, in December
1994, the Medical Devices Agency of the Department of Health in the United King-
dom reported their evaluation of the evidence for an association between breast im-
plants and connective-tissue diseases.^ The resulting 60-page document, which in-
cluded critical appraisals performed by an independent expert advisory group, con-
cluded that "there is no evidence of an increased risk of connective tissue diseases
in patients who have had silicone gel breast implants and therefore no scientific
case for changing practice or policy in the United Kingdom with respect to breast
implantation". Likewise the French Ministiy of Health stated in a press release on
January 24, 1995 that "an analysis of international scientific literature dem-
onstrates that the risk of a patient developing an autoimmune disease or cancer fol-
lowing the implantation of gel-filled breast prostheses is no greater than the risk
of such diseases in the general population . . . and that the moratorium of January