Cassie Part 3 Lymphocytes and Inflammatory Reaction to Silicone Gel Breast

[Guest guest]

http://cvi.asm.org/cgi/reprint/3/2/156.pdf,

CLINICAL AND DIAGNOSTIC LABORATORY IMMUNOLOGY, Mar. 1996, p. 156–161 Vol. 3, No. 2

1071-412X/96/$04.0010

Copyright q 1996, American Society for Microbiology

Phenotype of Lymphocytes Associated with the Inflammatory

Reaction to Silicone Gel Breast Implants

WILLIAM E. KATZIN,1,2* LU-JEAN FENG,2,3 MARY ABBUHL,1 AND MARY ANN KLEIN1

Department of Pathology1 and Division of Plastic Surgery,3 Mt. Sinai Medical Center, and Case Western Reserve

University School of Medicine,2 Cleveland, Ohio

Received 18 July 1995/Returned for modification 13 September 1995/Accepted 27 November 1995

The tissue response to silicone gel breast implants typically includes an inflammatory infiltrate that consists

of macrophages, foreign body-type giant cells, and a variable number of lymphocytes and plasma cells. The

phenotype of the lymphocytic component was investigated with three-color flow cytometry. Lymphocytes were

obtained by collecting fluid from the space between the implant and the fibrous capsule or by washing cells

from the fibrous capsule at the time of implant removal with total capsulectomy. Eighty-nine percent of the

implant-associated lymphocytes were T cells. Twenty-five percent of the CD31 T cells coexpressed HLA-DR

compared with only 7.9% of matched peripheral blood lymphocytes. Sixty-eight percent of the implantassociated

T cells coexpressed CD4 and CD29, while only 3% of the T cells coexpressed CD4 and CD45RO. The

expression of HLA-DR and the predominance of CD291 CD41 T cells indicate that there is immune activation

with the potential for stimulating antigen-specific antibody production. The role of silicone gel breast implants

in immune activation and its clinical significance require further investigation.

Silicone gel-containing breast implants, introduced over 30

years ago (7), have been widely used for augmentation and

reconstructive mammoplasty. Recently there has been increasing

concern regarding the safety of breast implants. Most notably,

a variety of rheumatic diseases, including chronic arthropathy

(2, 9, 39, 46), scleroderma or systemic sclerosis (6,

20, 21, 33, 36, 37, 40, 42), rheumatoid arthritis (9, 20, 40),

systemic lupus erythematosus (36, 40), and Sjogren’s syndrome

(20, 40) have been reported in patients with silicone gel breast

implants. However, the relationship between implants and

these connective tissue diseases is still unclear (41). Three

separate epidemiologic studies (11, 34, 47) have not found an

increased incidence of specific rheumatic diseases in large populations

of women with silicone gel implants. The possible

relationship between silicone breast implants and systemic disease

has been recently reviewed by -Guerrero et al.

(35).

The tissue response to silicone gel-containing breast implants

includes formation of a fibrous capsule. On the basis of

studies with an animal model, it is likely that the capsule forms

in a relatively short time—probably within 2 months (27). The

histologic features of the fibrous capsule in women with implants

have been described previously (3, 12, 14, 15, 30, 32, 38).

In addition to a band of dense fibrous tissue, the capsule

includes a variable number of inflammatory cells. Macrophages

with abundant vacuolated cytoplasm are a relatively

constant feature. There are often cyst-like spaces that contain

refractile, nonbirefringent clear material that almost certainly

represents some form of silicone (1, 14, 32, 38). The inflammatory

infiltrate also includes a variable number of lymphocytes

and plasma cells (12, 16, 38).

Aside from the morphologic features of the fibrous capsule

that surrounds silicone gel breast implants, very little is known

about the nature of the inflammatory response. In an attempt

to provide more specific information in this regard, we have

defined the phenotype of the lymphocytic component of the

capsular infiltrate by three-color flow cytometry.

MATERIALS AND METHODS

During the study period (1 May 1992 to 1 April 1993), a total of 209 patients

underwent implant removal. The vast majority of these patients presented with

local breast pain and systemic symptoms that they suspected were associated with

their implants. Surgery in all cases was performed by one of the authors (L.-J.F.),

who submitted all of the samples. All of the patients were examined by a

rheumatologist, internist, or neurologist prior to implant removal. Eight of the

209 patients had documented autoimmune diseases which had developed after

implantation. Three had rheumatoid arthritis, two had multiple sclerosis, one

had systemic lupus erythematosus, one had systemic lupus erythematosus and

polymyositis, and one had scleroderma. Only one of the 209 patients had evidence

of frank infection.

In preliminary studies, we found that a sufficient number of lymphocytes for

flow cytometric analysis could be obtained only from patients with either polyurethane-

coated silicone gel breast implants or textured-surface silicone gel

breast implants. In those patients, exudative fluid was present in the space

between the implant and the surrounding fibrous capsule. Little or no exudative

fluid or very few mechanically dislodged cells could be obtained from patients

with smooth-shell silicone gel implants. Implant-associated lymphocytes were

obtained at the time of surgical implant removal in all cases in which exudative

fluid was available. In some cases, fluid was collected from the space between the

implant and the surrounding fibrous capsule and sent directly for phenotyping. In

the other cases, the fibrous capsule was vigorously washed in RPMI 1640 medium

and the mechanically dislodged cells were immediately sent for phenotyping. In

a few cases, peri-implant fluid was sent from one breast, and a capsular wash was

sent from the opposite breast of the same patient. Implant-associated cells were

washed twice in RPMI 1640 medium, and the cell pellet was resuspended in 1 ml

of RPMI 1640 medium. The cell yield and viability were determined at this time.

The resuspended cells were then mixed with RPMI 1640 medium containing

10% fetal calf serum (Hyclone, Logan, Utah) and rocked in a 378C incubator for

30 min. Cell suspensions were filtered through a 40-mm-pore-size nylon mesh

(Tetco, Briar Cliffe Manor, N.Y.) and centrifuged for 5 min at 400 3 g. Cells

were then washed once in RPMI 1640 medium and resuspended in RPMI 1640

medium to obtain a cell concentration of 2 3 106 cells per ml. The cells in

samples with a low yield were resuspended in 1 ml of RPMI 1640 medium.

In addition to the implant-associated cells, heparin-anticoagulated peripheral

blood was submitted for phenotyping. Peripheral blood mononuclear cells were

isolated from heparinized blood by density gradient centrifugation (Histopaque-

1077; Sigma, St. Louis, Mo.). The peripheral blood mononuclear cells were

recovered and washed twice with Hanks balanced salt solution without Ca or Mg

(Gibco, Grand Island, N.Y.) at room temperature. The cell pellet was then

resuspended in 1 ml of Hanks balanced salt solution, and the cell count was

adjusted to 2 3 106 cells per ml for staining. In some cases, the lymphocytes were

separated from whole blood by Ficoll-Hypaque density gradient centrifugation

and stored in freezing medium (25% calf serum, 8% dimethyl sulfoxide, 67%

* Corresponding author. Mailing address: Department of Pathology,

Mt. Sinai Medical Center, One Mt. Sinai Dr., Cleveland, OH 44106.

Phone: (216) 421-4403. Fax: (216) 421-3964.

156

Downloaded from cvi.asm.org by on January 22, 2008

Dulbecco’s modified Eagle’s medium) in liquid nitrogen prior to flow cytometry.

Previously frozen peripheral blood mononuclear cells were snap thawed in a

378C water bath and washed twice with Dulbecco’s phosphate-buffered saline

(D-PBS) without Ca or Mg (Gibco). The cell pellet was resuspended in 1 ml of

D-PBS. Viability was checked with trypan blue (Sigma).

Three-color staining was performed with 20 ml of mouse monoclonal antibodies

(Table 1), at an optimum titer, directly conjugated with fluorescein isothiocynate

(FITC), phycoerythrin (PE [or RD1]), or peridinin chlorophyll protein

(PERCP). One hundred microliters of the cell suspension was added to the

appropriate tubes and mixed. The samples were then incubated at room temperature

out of direct light for 15 min with gentle vortexing at 5-min intervals.

Samples with gross erythrocyte contamination or whole-blood samples were

lysed by incubation with FACSLyse (Becton Dickinson, San , Calif.). Cells

were then washed twice with Hanks balanced salt solution with 1% bovine serum

albumin (BSA) (Sigma) and 0.1% sodium azide (Fisher, Pittsburgh, Pa.). The

cell pellet was then resuspended in 0.5 ml of D-PBS with 1% BSA and 1%

ultrapure electron microscopy-grade formaldehyde (Methanol Free; Polysciences,

Washington, Pa.) for fixation. Negative controls, consisting of isotypically

matched nonimmune mouse immunoglobulin, were used to position the

cursors that defined positive and negative cells. Positive controls consisted of

peripheral blood lymphocytes from healthy donors. Cells were analyzed on a

FACScan flow cytometer with an argon ion laser emitting at 498 nm (Becton

Dickinson). Gates for acquisition of data were set by light scatter characteristics

and were verified by back-gating with cells stained for CD45 and CD14. A total

of 10,000 gated events were acquired with LYSYS analysis software (Becton

Dickinson), yielding a percentage of total cells positive for each antigen. In the

few samples that contained too few cells to acquire 10,000 events, as many gated

cells as possible were acquired. Analysis of data was performed with PAINT-AGATE

software (Becton Dickinson).

Statistical comparisons between the phenotypes of the implant-associated lymphocytes

and those of the peripheral blood were analyzed with the paired t test.

Statistical comparisons between the phenotypes of the peripheral blood lymphocytes

of the patients and those of the healthy controls were analyzed with the

unpaired t test.

RESULTS

All cases accessioned between 1 May 1992 and 1 April 1993

for which cells were submitted for phenotypic analysis by flow

cytometry were included in this study, with the following exceptions.

In three cases, there was an insufficient number of

cells to provide complete and reliable phenotypic data. In four

additional cases, the percentage of events in the lymphocyte

gate that were CD451 and CD142 was less than 80%. In all

cases included in the study, that percentage was greater than

85%. In three cases, separate samples were submitted from the

right and left breasts. In those cases, there were no significant

phenotypic differences between the two samples, and only the

data from the sample submitted first were considered for statistical

calculations regarding the paired peripheral blood samples.

There were no phenotypic differences between exudative

fluid samples and capsular wash samples. In one case, the

peripheral blood lymphocytes that had been frozen were not

viable when the cells were thawed.

Clinical data regarding the 17 patients included in this study

are summarized in Table 2. The women ranged in age from 31

to 55 years, with a mean age of 40 years. The implants had been

in place for an average of 3.7 years (range, 1.1 to 9 years). All

patients presented with local complaints of either pain in the

breast or chest wall or capsular contracture. In addition, all of

the patients subjectively reported one or more constitutional

complaints such as arthritis, myalgias, or chronic fatigue. None

of the 17 patients in this study had well-documented specific

rheumatologic diseases, and none had evidence of infection.

Eleven patients had polyurethane-coated silicone gel breast

implants, and the remaining six had textured-surface silicone

gel-containing breast implants (Table 2). Sixteen of the 17

patients had bilateral breast implants.

The samples analyzed by flow cytometry contained an average

of 4.5 3 106 cells (range, 0.5 3 106 to 15 3 106 cells). The

volume of exudative fluid averaged around 1 to 3 ml. Stained

cytospin preparations from both the exudative fluids and capsular

washes revealed a predominance of lymphocytes (mean,

62%) and macrophages (mean, 37%), with only a few segmented

neutrophils (mean, 2%).

Flow cytometry. The immunophenotypic data regarding

both the implant-associated and peripheral blood lymphocytes

are summarized in Tables 3 and 4, and representative contour

plots are shown in Fig. 1 and 2. The vast majority of the

implant-associated lymphocytes were T cells (mean, 89%;

range, 81 to 97%). The mean percentage of CD191 B cells was

only 1.4% (range, 0.0 to 4.6%). A small percentage (mean,

4.1%; range, 0.0 to 16%) of the implant-associated lymphocytes

were CD16/561 CD32 natural killer cells. Compared

with the paired peripheral blood lymphocytes, the implantassociated

lymphocyte population had a significantly greater

proportion of T cells (P , 0.001) and significantly fewer B cells

(P , 0.001) and natural killer cells (P , 0.001). In order to

compensate for this increased proportion of T cells, data regarding

T-cell subsets were normalized on the basis of the total

percentage of CD31 cells in order to provide a meaningful

TABLE 1. Antibody panel used in this study

CD (clone) Major specificity group Fluorescent

label

Dilution

Sourcea

CD45 (Hle-1) Pan-leukocyte FITC 1:1 BD

CD14 (Leu-M3) Monocytes PE 1:1 BD

CD3 (Leu-4) T cells FITC or

PERCP

1:1 BD

CD4 (Leu-3) T-helper and/or inducer

cells, monocytes

FITC or

PERCP

1:1 BD

CD8 (Leu-2) T-cytotoxic and/or

suppressor cells

FITC 1:1 BD

CD16/56 (Leu-

11c 1 Leu-19)b

Natural killer cells PE 1:1 BD

CD19 (Leu-12) B cells PERCP 1:1 BD

CD20 (Leu-16) B cells PERCP 1:1 BD

CD29 (4B4) T-cell subset RD1 1:5 CI

CD45RO (2H4) T-cell subset RD1 1:64 CI

HLA-DR PERCP 1:1 BD

a BD, Becton Dickinson Immunocytometry Systems, San , Calif.; CI,

Coulter Immunodiagnostics, Hialeah, Fla.

b In a few cases, CD56 (Leu-19) was used alone.

TABLE 2. Clinical data for the patients used in this study

Age

(yr) Implant typea No. of yr

with implant

Reason for

implantb

36 Meme (PU) 9 Rec

41 Replicon (PU) 6 Rec

35 Meme and Replicon (PU) 1.5 Rec, Aug

35 Meme (PU) 2.4 Aug

36 Meme (PU) 6 Aug

55 Optimam (PU) 8 Rec

41 Replicon (PU) 3 Aug

46 Replicon (PU) 2.3 Aug

34 Meme (PU) 7 Aug

31 MSI (TS) 1.3 Rec

45 Biocell (TS) 2.5 Aug

40 Biocell (TS) 2 Rec

47 Biocell (TS) 1.1 Aug

51 Optimam and Meme (PU) 2.3 Rec

37 Unknown (PU) 3.3 Aug

39 Misty (TS) 1.4 Rec

32 Mentor Siltex (TS) 2.5 Aug

a PU, polyurethane-coated silicone gel implant; TS, textured-surface Silastic

implant.

b Rec, reconstruction; Aug, augmentation.

VOL. 3, 1996 LYMPHOCYTES ASSOCIATED WITH SILICONE GEL IMPLANTS 157

Downloaded from cvi.asm.org by on January 22, 2008

comparison of these cell types between the two compartments.

Among the implant-associated T cells, there was increased

expression of HLA-DR (P , 0.001). Twenty-five percent of the

CD31 cells coexpressed HLA-DR. There was a predominance

of CD41 T cells; for the implant-associated cells, the mean

CD4/CD8 ratio was 1.8. Essentially all (greater than 99%) of

the CD4 cells coexpressed CD29. Furthermore, almost all of

the CD41 cells were negative for CD45RO. Compared with

the paired peripheral blood lymphocytes, the increase in CD41

CD291 cells was marginally significant (P 5 0.043) and the

decrease in CD41 CD45RO1 cells was statistically significant

(P , 0.001).

The peripheral blood lymphocyte subsets in the patients with

implants were, in most respects, similar to those in healthy

controls (Table 5). These data are only preliminary, since the

patient and control populations were not matched with regard

to sex or age. Nevertheless, the data do suggest that there is a

decrease in the number of CD41 CD45RO1 cells in patients

with implants (P 5 0.022). Other differences in lymphocyte

subsets did not appear to be statistically significant.

DISCUSSION

Despite concern regarding possible immunologic abnormalities

associated with silicone gel breast implants, the phenotype

of implant-associated lymphocytes has not previously been

clearly defined. In preliminary studies, others have noted a

predominance of T cells (29, 31). In this study, three-color flow

cytometry was used to define the phenotypes of the lympho-

FIG. 1. Representative contour plot of implant-associated lymphocytes stained with antibodies to CD4 and CD29 (4B4). The vast majority of the cells coexpress

CD4 and CD29 (upper right quadrant).

TABLE 3. Phenotype of implant-associated lymphocytes compared

with that of paired peripheral blood lymphocytes

Phenotype

% of lymphocytes positivea

Implant Peripheral blood

CD31 (T cells) 89 6 4.9 71 6 9.1

CD19 or CD201 (B cells)b 0.9 6 1.3 12 6 8.3

CD16/561 CD32 (natural killer cells) 3.6 6 3.8 15 6 7.2

a Values are means 6 standard deviations (n 5 16). For two-tailed probability,

P , 0.001 (paired t test).

b The B-lymphocytes in five of the peripheral blood samples were quantitated

with antibodies to CD20 instead of CD19.

TABLE 4. Implant-associated T-cell subsets compared with paired

peripheral blood T-cell subsets

Phenotype

% of T cells positivea

2-Tailed

probabilityb

Implant Peripheral blood

HLA-DR1 25 6 8.1 7.9 6 3.8 ,0.001

CD41 CD291 68 6 12 59 6 13 0.043

CD41 CD292 0.6 6 0.1 3.6 6 3.5 0.007

CD41 CD45RO1 2.6 6 5.2 19 6 7.2 ,0.001

CD41 CD45RO2 63 6 13 41 6 8.5 ,0.001

CD81 37 6 10 42 6 9.6 0.115

a Values are means 6 standard deviations (n 5 16).

b Paired t test.

158 KATZIN ET AL. CLIN. DIAGN. LAB. IMMUNOL.

Downloaded from cvi.asm.org by on January 22, 2008

cytes that are associated with polyurethane-coated and textured-

surface silicone gel-containing breast implants. These

types of implants were specifically chosen for this study because

of the large number of lymphocytes that could be harvested

from the capsule or from the peri-implant fluid.

Whether or not the inclusion of only textured-surface and

polyurethane-coated implants in our study introduced some

bias cannot be determined from our available data. In all cases,

there was a striking predominance of T cells. Most of the T

cells had the phenotype CD31 CD41 CD291 CD45RO2. Furthermore,

among the T cells there was significant expression of

HLA-DR. Comparison of the implant-associated lymphocytes

with patient-matched peripheral blood lymphocytes, obtained

at the time of surgery, showed that the increase in CD31 cells,

the increased expression of HLA-DR by the T cells, and the

decrease in expression of CD45RO by the CD41 cells were all

statistically significant (P , 0.001 for each comparison). The

increase in the percentage of CD41 CD291 T cells among the

implant-associated lymphocytes compared with the level of the

matched peripheral blood lymphocytes was of borderline statistical

significance (P 5 0.043).

The role of textured-surface implants in eliciting an exudative

host response is uncertain. It is known that textured-surface

Silastic implants, as well as polyurethane-covered implants,

are frequently associated with synovial metaplasia of

the lining of the fibrous capsule (30). Review of histologic

sections from the capsules from each of the patients in this

study confirmed the presence of synovial metaplasia in all

cases. In addition, in most capsules there was a marked lymphocytic

infiltrate together with foamy macrophages.

The significance of the predominance of CD291 CD41 T

cells among the implant-associated lymphocytes is uncertain.

CD41 T cells can generally be divided into two major functional

categories. CD41 CD291 T cells proliferate maximally

to soluble antigen and increase antigen-specific antibody production

(22). In contrast, CD41 CD45RO1 T cells induce CD8

FIG. 2. Representative contour plot of peripheral blood lymphocytes stained with antibodies to CD4 and CD45RO (2H4). A significant fraction of the CD41 cells

coexpress CD45RO (upper right quadrant). PerCP, PERCP.

TABLE 5. Patient peripheral blood lymphocytes compared with

control peripheral lymphocytes

Phenotype

% of lymphocytes positivea

2-Tailed

Implant patients probabilityb

(n 5 16)

Controls

(n 5 12)

CD3 71 6 9.1 75 6 3.5 0.072

CD31 HLA-DR1 5.5 6 2.5 3.9 6 2.3 0.097

CD41 CD291 41 6 10 45 6 9.2 0.301

CD41 CD45RO1 13 6 5.5 21 6 9.8 0.022

CD81 30 6 7.6 25 6 7.1 0.081

CD191 (or CD201)c 12 6 8.3 15 6 4.1 0.186

CD16/561 CD32 15 6 7.2 11 6 5.0 0.069

a Values are means 6 standard deviations.

b Unpaired t test.

c B lymphocytes in five of the patient peripheral blood samples were quantitated

with antibodies to CD20 instead of CD19.

VOL. 3, 1996 LYMPHOCYTES ASSOCIATED WITH SILICONE GEL IMPLANTS 159

Downloaded from cvi.asm.org by on January 22, 2008

cells to exert suppressor function (23). It is of interest that

CD41 CD291 T cells are also the predominant lymphocyte

subset in synovial fluid of patients with rheumatoid arthritis

(24). Furthermore, as we found for implant-associated lymphocytes,

synovial tissue lymphocytes in patients with rheumatoid

arthritis have increased expression of HLA-DR antigens (8).

The apparent depletion of CD41 CD45RO1 T cells from the

peripheral blood of patients with silicone gel-containing breast

implants will require confirmation by more extensive controlled

studies.

To date, there is very little information that sheds light on

the manner whereby silicone gel-containing breast implants

could potentially cause connective tissue diseases. Some authors

have suggested that silicone or silicone-protein complexes

may themselves be antigenic (13, 18, 19, 44). Others

have drawn attention to the fact that a significant percentage of

women with silicone breast implants have antinuclear antibodies

(5, 28). The results from animal studies have been conflicting

(4, 25). Clinical studies of patients with silicone gel breast

implants suffer from the fact that patients are often referred

because of symptoms of rheumatic disease. Large populationbased

epidemiologic studies are limited by the fact that ‘‘diseases’’

potentially caused by silicone gel-containing implants

have not been clearly defined. Most of the patients in this study

reported a symptom complex similar to fibromyalgia. A preliminary

study of 144 patients with breast implants also reported

clinical manifestations characteristic of fibromyalgia

(45). Certainly there is no reproducibly specific marker for

patients with silicone gel-associated rheumatic disease. In one

study, which has yet to be substantiated, Vasey and colleagues

reported clinical observations suggesting that rheumatic disease

symptoms in patients with silicone gel implants may be

reversible after implant removal (43). This finding should at

least provide the impetus for continued studies of the immunologic

effects of silicone.

There are several inherent limitations in our study regarding

its ability to address the questionable relationship between

silicone breast implants and immunologic diseases. First, our

patient population includes only those patients who are currently

having problems with their implants. Because it is difficult

to form a control group of patients with no local or systemic

problems and who wish to undergo implant removal and

capsulectomy, this study does not resolve the question of

whether T-cell activation around polyurethane or texturedsurface

breast implants is in fact a pathologic phenomenon.

Although polyurethane-covered implants have been known to

elicit greater inflammatory responses in the capsule than

smooth-shell implants, it is not clear from previous studies

what type of inflammation they elicit (16). Nor is it clear from

previous studies how long the inflammatory responses persist.

Our study has shown that T-cell activation around these implants

occurs as early as 1 year after implantation and can

persist for as long as 9 years.

The other unresolved question which this study could not

address is the significance of local T-cell activation in eliciting

local pain and systemic illness. Although all patients studied

had significant pain around their implants and all of the patients

had some concomitant constitutional symptoms, without

a large symptom-free control group, the effect of local T-cell

activation on the systemic immune system remains elusive.

In summary, this study is the first to define the phenotype of

the lymphocytes isolated from the fluid or tissue surrounding

silicone gel-containing breast implants. The lymphocytes are

almost all CD31 T cells, most of which express CD4 and CD29.

Compared with peripheral blood lymphocytes, the T cells have

significantly increased expression of HLA-DR and significantly

reduced expression of CD45RO. Histologic examination of

capsular tissue, as well as examination of the exudative fluid,

has shown that most of the implant-associated lymphocytes are

present in association with foamy macrophages that are known

to ingest droplets of silicone gel. We have found that these

macrophages strongly and uniformly express HLA-DR (17).

Our results are consistent with the hypothesis that the siliconecontaining

macrophages act as antigen-presenting cells to CD4

cells which become activated and subsequently function to

up-regulate an immune response, as indicated by expression of

CD29. Admittedly, this simplistic hypothesis requires extensive

further testing and refinement. In particular, it will be important

to determine whether the implant-associated CD41 T

cells belong to the Th1 or Th2 subset. A recent study by

Ojo-Amaize et al. (26) provides additional evidence in support

of the central role of T cells in the immunologic reaction to

silicone breast implants. In any case, if science is to prevail in

the silicone controversy, as Fisher rightly insists (10), then

further analysis of the inflammatory reaction to silicone breast

implants will be necessary.

ACKNOWLEDGMENTS

This research was supported in part by a grant from the Haas Fund,

The Mount Sinai Medical Center, Cleveland, Ohio.

We express our gratitude to Hom of the Department of

Epidemiology and Biostatistics at Case Western Reserve University

School of Medicine, Cleveland, Ohio, for assistance with the statistical

analyses.

REFERENCES

1. Baker, J. L., R. R. LeVier, and D. E. Spielvogel. 1982. Positive identification

of silicone in human mammary capsular tissue. Plast. Reconstr. Surg. 69:56–

60.

2. Baldwin, C. M., and E. N. Kaplan. 1983. Silicone-induced human adjuvant

disease? Ann. Plast. Surg. 10:270–273.

3. Brandt, B., V. Breiting, L. Christensen, M. Nielsen, and J. L. Thomsen. 1984.

Five years experience of breast augmentation using silicone gel prostheses

with emphasis on capsule shrinkage. Scand. J. Plast. Reconstr. Surg. 18:311–

316.

4. Brantley, S. K., S. F. son, P. A. St. Arnold, M. B. , P. J. Talbot,

J. B. Grogan, M. A. Cuchens, H. S. Hsu, and S. K. Das. 1990. Assessment of

the lymphocyte response to silicone. Plast. Reconstr. Surg. 86:1131–1137.

5. Bridges, A. J., C. Conley, G. Wang, D. E. Burns, and F. B. Vasey. 1993. A

clinical and immunologic evaluation of women with silicone breast implants

and symptoms of rheumatic disease. Ann. Intern. Med. 118:929–936.

6. Brozena, S. J., N. A. Fenske, C. W. Cruse, C. G. Espinoza, F. B. Vasey, B. F.

Germain, and L. R. Espinoza. 1988. Human adjuvant disease following

augmentation mammoplasty. Arch. Dermatol. 124:1383–1386.

7. Cronin, T. D., and F. J. Gerow. 1964. Augmentation mammoplasty: a new

‘‘natural feel’’ prosthesis, p. 41. In Transactions of the Third International

Congress of Plastic Surgery. Excerpta Medica, Amsterdam.

8. Cush, J. J., and P. E. Lipsky. 1988. Phenotypic analysis of synovial tissue and

peripheral blood lymphocytes isolated from patients with rheumatoid arthritis.

Arthritis Rheum. 31:1230–1238.

9. Endo, L. P., N. L. , S. Longley, L. C. Corman, and R. S. Panush.

1987. Silicone and rheumatic diseases. Semin. Arthritis Rheum. 17:112–118.

10. Fisher, J. C. 1992. The silicone controversy: when will science prevail? N.

Engl. J. Med. 326:1696–1698.

11. , S. E., W. M. O’Fallon, L. T. Kurland, C. M. Beard, J. E. Woods, and

L. J. Melton. 1994. Risk of connective-tissue diseases and other disorders

after breast implantation. N. Engl. J. Med. 330:1697–1702.

12. Gayou, R. M. 1979. A histological comparison of contracted and non-contracted

capsules around silicone breast implants. Plast. Reconstr. Surg. 63:

700–707.

13. Goldblum, R. M., R. P. Pelley, A. A. O’Donell, D. Pyron, and J. P. Heggers.

1992. Antibodies to silicone elastomers and reactions to ventriculoperitoneal

shunts. Lancet 340:510–513.

14. Hardt, N. S., L. T. Yu, G. La Torre, and B. Steinbach. 1994. Fourier transform

infared microspectroscopy used to identify foreign materials related to

breast implants. Mod. Pathol. 7:669–676.

15. Hester, T. R., F. Nahai, J. Bostwick, and J. Cukic. 1988. A 5-year experience

with polyurethane-covered mammary prostheses for treatment of capsular

contracture, primary augmentation, mammoplasty, and breast reconstruction.

Clin. Plast. Surg. 15:569–585.

16. Jabaley, M. E., and S. K. Das. 1986. Late breast pain following reconstruc-

160 KATZIN ET AL. CLIN. DIAGN. LAB. IMMUNOL.

Downloaded from cvi.asm.org by on January 22, 2008

tion with polyurethane-covered implants. Plast. Reconstr. Surg. 78:390–395.

17. Katzin, W. E., and L. J. Feng. 1994. Phenotype of lymphocytes associated

with the inflammatory reaction to silicone-gel breast implants. Mod. Pathol.

7:17A. (Abstract.)

18. Kossovsky, N., and C. J. Freiman. 1994. Silicone breast implant pathology:

clinical data and immunologic consequences. Arch. Pathol. Lab. Med. 118:

686–693.

19. Kossovsky, N., J. P. Heggers, and M. C. Robson. 1987. Experimental demonstration

of the immunogenicity of silicone-protein complexes. J. Biomed.

Mater. Res. 21:1125–1133.

20. Kumagai, Y., A. Chiyuki, and Y. Shiokawa. 1979. Scleroderma after cosmetic

surgery: four cases of human adjuvant disease. Arthritis Rheum. 22:532–537.

21. Marik, P. E., A. L. Kark, and A. Zambakides. 1990. Scleroderma after

silicone augmentation mammoplasty: a report of 2 cases. S. Afr. Med. J.

77:212–213.

22. Morimoto, C., N. L. Letvin, A. W. Boyd, M. Hagan, H. M. Brown, M. M.

Kornacki, and S. F. Schlossman. 1985. The isolation and characterization of

the human helper inducer T cell subset. J. Immunol. 134:3762–3769.

23. Morimoto, C., N. L. Letvin, J. A. Distaso, H. M. Brown, and S. F. Schlossman.

1986. The cellular basis for the induction of antigen-specific T8 suppressor

cells. Eur. J. Immunol. 16:198–204.

24. Morimoto, C., P. L. Romain, D. A. Fox, P. , M. DiMaggio, H.

Levine, and S. F. Schlossman. 1988. Abnormalities in CD41 T-lymphocyte

subsets in inflammatory rheumatic diseases. Am. J. Med. 84:817–825.

25. Naim, J. O., R. J. Lanzafae, and C. J. van Oss. 1993. The adjuvant effect of

silicone-gel on antibody formation in rats. Immunol. Invest. 22:151–161.

26. Ojo-Amaize, E. A., V. Conte, H.-C. Lin, R. F. Brucker, M. S. Agopian, and

J. B. . 1994. Silicone-specific blood lymphocyte response in women with

silicone breast implants. Clin. Diagn. Lab. Immunol. 1:689–695.

27. Picha, G. J., and J. A. Goldstein. 1990. Analysis of the soft-tissue response to

components used in the manufacture of breast implants: rat animal model.

Plast. Reconstr. Surg. 87:490–500.

28. Press, R. I., C. L. Peebles, Y. Kumagai, R. L. Ochs, and E. M. Tan. 1992.

Antinuclear autoantibodies in women with silicone breast implants. Lancet

340:1304–1307.

29. Raso, D. S. 1994. Breast prostheses, the immune response, and B- and

T-lymphocytes. Plast. Reconstr. Surg. 93:649–650.

30. Raso, D. S., L. W. Crymes, and J. S. Metcalf. 1994. Histological assessment

of fifty breast capsules from smooth and textured augmentation and reconstruction

mammoplasty prostheses with emphasis on the role of synovial

metaplasia. Mod. Pathol. 7:310–316.

31. Rowlands, C., and F. H. Y. Green. 1994. Tissue responses to silicone breast

implants. Mod. Path. 7:21A. (Abstract.)

32. Rudolph, R., J. Abraham, T. Vecchione, S. Guber, and M. Woodward. 1978.

**************One site keeps you connected to all your email: AOL Mail, Gmail, and Mail. Try it now. (http://www.aol.com/?optin=new-dp & icid=aolcom40vanity & ncid=emlcntaolcom00000025)