SIRG minutes - 14 May 2001

Silicone Gel Breast Implants: Independent Review Group Meeting
14 May 2001, Hannibal House, Elephant and Castle

Present:
     Professor R Sturrock (Chairman)
     Mrs V Harpwood
     Professor D London
     Mr T Milward
     Professor R Walker
     Ms N Ratcliff

Apologies:
     Professor R Batchelor
     Professor A Silman
     Ms K Murphy

Secretariat:
     Dr A Austin
     Dr S Ludgate
     Mrs A Miller
     Mr J Tinkler

In attendance:
     Dr P Troop
     Dr D Jeffereys

1. Introduction
The Chairman welcomed all members of the Group present.

2 & 3 Minutes of the last meeting
The minutes of the previous meeting held on the 14 December 2000 (SIRG(01)01) were confirmed. There were no matters arising from these minutes which were not already on the agenda of this meeting.

4. Update on the recommendations
4.1. Recommendations 1. The 80,000 patient information booklets which were originally printed have all been sent out. The document is being revised prior to reprinting. The Department of Health is at present considering whether to reconvene the Emerton Group which originally drafted the document.

4.2 Recommendation 2. The Care Standards Act, which came info force in 2000, requires that advertisements and literature relating to private clinics are not misleading. The Health Select Committee (HSC) which looked at breast implants in March of this year also raised concern regarding advertising.

4.3. Recommendation 3. Clinical Guidelines for surgeons carrying out breast implant surgery are being drafted by a working party convened by the Royal College of Surgeons under the chairmanship of Mr Brough. The IRG have asked that they be given an opportunity to comment on the draft document prior to publication.

4.4. Recommendation 4. Mr Brough's Group is considering if a comprehensive surgery consent form as part of the Guidelines discussed in 4.3.

4.5. Recommendation 5. The Health and Social Care Act was drafted to ensure that private clinics and hospitals provide the same standard of care as the NHS . The IRG was influential identifying the need for this legislation. A Commission has been put in place by secondary regulation to police the Act including the inspection of private hospitals.

4.6 Recommendation 6. It was noted that the HSC also recommended that registration on the NBIR be compulsory. Although this is not possible a Steering Group was established in 1999 to advise on strategies to ensure optimal compliance.

4.7. Recommendation 7. MDA issued guidance on the reporting of adverse incidents under the Vigilance system in August 1999. This guidance document was distributed to surgeons though the professional bodies. In addition to this, MDA is in the early stages of drafting another document on adverse incident involving breast implants aimed directly at the surgeons. The adverse incident system is designed to detect problems within particular models of breast implants. The NBIR data collection form has been amended to include a section to gather data on the reason an implant was removed. It is felt that this will help in monitoring the lifetime of implants

4.8. Recommendation 8. A pilot study has been commissioned by the Department of Health, Research and Development Directorate to investigate the feasibility of using the data held on the NBIR and to look at early complications of breast implant surgery. The group was concerned that its recommendation to study chronic low grade infection had not been taken forward. The chair agreed to write to the Chief Medical Officer and the head of the Research and Development Directorate to raise this issue. Immunological research was also not being taken forward activly although the group continues to review published data.

4.9 Recommendation 9. The group continues to monitor any new scientific papers in this area.

The chairman was encouraged that much has happened overall in meeting the recommendations and noted that the fact that the IRG continued to meet was a major force in ensuring that the recommendations were being taken forward.

5. Membership of the IRG
5.1 The chairman noted that Claire Rayner had resigned from the Group at the end of last year. He had contacted both the Consumers Association and the patient's Association for recommendations and two new Lay Members had been appointed These were Nikki Ratcliff (Consumers Association) and Catherine Murphy (Patients Association).
The Chairman welcomed the new member present and discussed the remit of the IRG. He stressed that the IRG was independent and that although MDA acted as secretariat and sat in on meetings, they did not participate in discussions other than to provide factual information when requested to do so. If sensitive issues are to be discussed MDA leaves the meeting.

5.2. At the previous meeting it had been agreed that an immunotoxicologist should be identified and the British Toxicology Society was approached to recommend suitable a candidate. The Group reiterated that members should not be employed by or funded by any industrial organization and all possible conflicts of interests should be declared to the Group.

5.3. The Chairman informed the Group that Richard Batchelor had tended his resignation from the Group for personal reasons. The Group noted the important part he had in the work of the IRG and thanked him for his hard work in this area, noting that his ability to explain complex immunological issues to non-immunologists would be especially missed.

6.IRG Website
It was noted that the current website service provider would no longer be able to maintain the IRG's website due to changes in the department of Heath's website arrangements. The Group noted that a new provider would be identified by the secretariat in line with DoH recommendations. It was noted that the appearance of the website might alter and that, until a new provider was in place, there might be a delay in updating the website.

7. Breast Implants and Mammography
At the previous meeting the need to update the website to give some information regarding the effects of breast implants on mamography had been raised. A draft document was being prepared which would include information on the effect that any calcification may have on interpretation and additional investigations. The Group requested that the updated Information Booklet should include information on this subject.

8. Breast Implants and Breast Feeding
Dr Zuckerman had supplied data relating to the effects of breast implants on breast feeding. The Group noted that problems were particularly evident in women whose implants had been inserted in the perialveolar area, as this technique cut the milk ducts. This method of insertion is believed to be rarely used in the UK. The Group requested that the updated Information Booklet should include information on this issue.

9. Health Select Committee hearing on breast implants
The Chairman informed the Group that he had recently represented the Group at a Health Select Committee (HSC) hearing on breast implants. It was noted that many of the recommendations of the HSC were similar to those of the IRG, particularly in relation to concerns about advertising, private clinics, informed consent, the registry and adverse incident investigations. The HSC did not question the findings or recommendations of the IRG.

The Group noted that the HSC was particularly concerned about the information women received prior to surgery and recommended that they see an independent councilor.

10. Platinum toxicity
A paper was presented to the Group and it was agreed that further work was needed to review the chemical composition of platinum compounds used in the manufacturer of breast implants, the stability of these compounds and exposure to them.

11. Scientific papers
Member present summarized their findings from the papers they had reviewed, which included recently published or submitted but unpublished scientific papers. Details of the reviews of published papers can be found in the annex to these minutes. Members present agreed that these papers did not alter the IRG's original conclusions, but one or two papers had identified areas that required continued monitoring.

12. Action in Europe
It was noted that the European Commission's "Communication" regarding breast implants had not yet been presented to the European Parliament. If approved Member States would not be invited to implement the measures contained in the document. It was noted that the UK already complied with the majority of recommendations in the most recent draft.

13. Correspondence from interested parties
13.1 Professor Shanklin had submitted several reports which either referenced published articles or gave details of unpublished work. The data, which had not been previously reviewed by the IRG, was included in the scientific paper review which was discussed at item 11and summarized in the Annex to these minutes.

13.2 Dr Zuckerman had submitted several reports which referenced published articles. The data which had not been previously reviewed by the IRG was included in the scientific paper review which was discussed at item 11and summarized in the Annex to these minutes. Dr Zuckerman had raised the issue of the effects of breast implants on the ability to breast feed-which was discussed under item 8.

13.3 Letters were also received from members of the public: Theses relate to the Open Day and the need to respond points raised by individual women and evidence provided by individual clinicians/scientist. The Group noted that it was more than happy to receive any points of information from individual members of the public, but could not continue to be involved in lengthy and detailed correspondence on matters which had already been discussed in the past.
Letters from the IRG had been sent to individual clinicians/scientist referred to during the Open Day; to date not all had replied. The Group agreed to review the information one available supplied and make details of this available.

14. AOB
14.1 Pat Troop, Deputy Cheif Medical Officer attended part of the meeting to express her thanks to the members of the Group. She stressed the importance of the work which has been carried out by the Group and highlighted the need for this to continue.

14.2. The Chairman noted that Avril Miller (who had given administrative support to the IRG) was retiring. The Group thanked her for her help and hard work.

May 2001: Annex 1 : Recently published literature : Summaries provided by IRG members

Raso D et.al., "Elemental analysis and clinical implications of calcification deposits associated with silicone breast implants" Ann Plast Surg 1999 Feb 42:2, 117-23 [SIRG(001)26]
This study involved the analysis of 29 deposits, from 26 patients, of silicone breast implant-associated calcifications, using electron probe microanalysis and infra red spectroscopy. Calcifications that occur following implantation could lead to problems in interpretation of mammograms and have an impact on screening. Analysis showed that calcium was complexed with tribasic phosphate, but there was no evidence of calcium oxalate, calcium carbonate, silicone or talc. The calcium phosphate may therefore cause problems in mammographic interpretation, since there could be problems in distinguishing whether it related to an implant or a carcinoma.

This paper describes the chemical composition of the pericapsular calcification observed around silicone implants in specimens which have been removed at explant surgery. The authors make the point that the calcification consists of Calcium Phosphate and that no silicone was found in the calicified areas of breast tissue. Breast tumours are associated with calcification and therefore baseline mammography should be performed on women who have had explant surgery in order to avoid over interpreting the significance of breast tissue calcification in explanted women who undergo routine screening mammography.

Eliopoulos DA and Lyle G, "Mycobacterium avium infection in a patient with the acquired immunodeficiency syndrome and silicone breast implants" South Med J 1999 Jan 92:80-3 [SIRG(001)27]
This is a single case report of atypical mycobacterial infection in an implanted breast occurring in a woman with AIDS.
Minor trauma to the affected breast had been sustained prior to the infection and the infected implant had ruptured. The report highlights the fact that opportunistic infection can occur in immunocompromised patients who have had a breast implant.

James SE et.al., "Silicone in the sputum after rupture of a calf implant" Journal of the Royal Society of Medicine March 2001 vol 94 133-134 [SIRG(001)28]
This is a single case report of putative venous and pulmonary embolism of silicone gel from a ruptured calf implant.
Case report describing a patient who received bilateral silicone calf implants 5 years previously and who experienced right calf swelling and pain. She also had lethargy, gastritis and backache and coughed up a lumpy, sticky substance. Ultrasound showed leakage from the R. implant, and at removal both were shown to be ruptured. Electron microscopy of sputum revealed deposits morphologically corresponding to silicone oxide and analysis showed it to match that from the implants. The mechanism of migration from the calf to the lungs is unclear but the authors speculate that it could have embolized through the vascular system.

Shanklin DR and Smalley DL, "Scleroderma in children of women with antenatal silicone devices". Southern Society for paediatric research/American federation for medical research, southern regional meeting, New Orleans Feb 18 1999 J Invest Med 1999 47:107A [SIRG(001)19]
This is an abstract reporting the occurrence of Scleroderma in 2 families of women with a breast implant. No details are provided of the family histories of the women involved and not enough information is available to properly assess this report.

Flassbeck D, et.al., Determination of low molecular weight silicones in plasma and blood of women after exposure to silicone breast implants by GC/MS. Anal Chem 2001, 73, 606-611 [SIRG(001)29]
The goal of the study was to develop a very sensitive, one-step sample preparation method for the detection of low molecular weight cyclic silicones in human plasma and blood. The authors used gas chromatography/mass spectroscopy to measure low molecular weight silicones in the blood and plasma of a small number of women who had had a breast implant and compared the levels in a control group of women. In initial experiments, blood and plasma were spiked with cyclic siloxanes (D3-6). Extraction efficiencies were 90% for plasma and 80% for blood but recovery for D3 was very low.

Ten plasma samples (stored at 3°C for nearly one year) and blood samples (stored at -20°C for three months) were analysed from women with current or previous silicone gel breast implants. Controls were six plasma and two blood samples stored the same ways. No disease controls e.g. insulin dependent diabetics were studied. Each sample was analysed 3 times. Cyclic siloxanes were detected in the implanted women but not in the healthy controls or in women with ruptured implants. D3 and D4 could be detected in 7 0f 10 plasma samples and 2 of 4 blood samples from women who had implants in situ or who had had implants for 22 years and had them removed 4 years prior to blood sampling. There was no correlation between silicone levels and duration of implant or the presence of bleed or rupture. The calculated plasma and blood silicone values were in the lower range of accepted total "normal" silicon.

The clinical data were difficult to determine and the very small number of individuals studied makes interpretation of the data difficult. Further studies using this methodology would be needed to establish the significance of the results.

Brown, L et.al., Silicone gel breast implant rupture, extracapsular silicone and health status in a population of women". J Rheumatol 2001;28:996-103 [SIRG(01)39]
This paper is still under review by the IRG.

Hovi S-L et.al., "Cosmetic and postmastectomy breast implants: Finnish women's experiences" J Women Health Gend Based Med 1999 Sep 8:7, 933-9 [SIRG(001)30]
In this study 224 women were recruited through advertising in the mass media with a 91% response rate for a questionnaire asking them their experience both positive and negative with silicone breast implants. 113 responders had their implants for breast reconstruction and 111 had them for cosmetic augmentation.
58% of women (equal between the two groups) said they had insufficient knowledge of breast implants pre-operatively.
26% of women said they would not choose the implants again.
44% of women expressed no dissatisfaction with their breasts (presumably meaning they were satisfied with the final result).
The authors conclude that because the insertion of breast implants is a life long decision that in depth counselling about complication rates and possible risks should be given before implantation and non-surgical alternatives should be discussed. (This is very much in line with the Independent Review Group's advice.)

Nietert PJ and Silver RM, Curr Opin Rheumatol 2000 Nov;12(6):520-6
Systemic sclerosis environmental and occupational risk factors [SIRG(001)18]
The article reviewed the current medical research that has examined associations between environmental agents and the development of systemic sclerosis (SSc). Very little attention was given to silicone implants.

Results:

1. Environmental agents and scleroderma-like disorders
2. • PVC production (US/Germany) - reports of development of Raynaud's phenomenon and SSc for people involved in the production of vinyl monomer.
   • Toxic oil syndrome (Spain) - 22% patients had SSc-like skin changes.
   • L-tryptophan/Eosinophilia myalgia syndrome.
3. Silica
4. • 1914 - stonemasons and silica dust was the first study and there have been many since.
   • East Germany study (1981-88) - men exposed to silica were 25x more likely to develop SSc - this is now an occupational disease in Germany, South Africa and Canada.
   • However, some studies have reported no association.
5. Silicone breast implants
6. • Most studies have found no association between implants and SSc. The authors refer to a large case-control in Michigan which found no increased risk (OR=1.30, 95% CI 0.27-6.23) and to a meta-analysis of 7 case-control studies which reached a similar conclusion (OR=1.03, 95% CI 0.60-1.62.
7. Organic solvents
8. • First report in 1957 - there have been 6 case-control studies that assessed solvent exposure since. Also at least 4 other cases have been published in abstract/letter form. Also one retrospective cohort study (Lundberg et al. 1994)
   • High potential for recollection bias in these cases.
   • Most report that patients with SSc are more likely to have been exposed to organic solvents/epoxy resins/silicates/paint thinners or removers.
   • Increased odds of exposure to organic solvents in those patients with SSc and who were positive for anti-SCL-7 compared to those with SSc who were negative (Nietert et al. 1999) linked to solvent-orientated hobbies.
   • Lacey et al. 1999 reported a link between petroleum distillate solvent exposure (through various occupations) and undifferentiated connective tissue disease in women (Michigan/Ohio). Also published reporting a link with SSc (abstract 1998).
9. Other environmental factors
10. • Expoxy resins: mainly case reports but several case-control studies none with significant results.
    • Hand/arm vibration: mainly as this appears to be a risk factor in Raynaud's phenomenon, however the association is unclear. A Japanese series reported high risk but no controls were used. Of the case-control studies only 1 assessed if this was a significant risk factor for SSC and they concluded it was not (OR=1.36, 95% CI 0.15-9.88).
    • Pesticides

Discussion:
Although there appears to be evidence that environmental agents play a role in SSc development the exact role is unclear.
Further study would be aided by following the steps reported by (Miller et al. Arth Rheum 2000; 43: 243-249) for reporting environmentally associated rheumatic disorders (EARD)

Report case/series that meet criteria for EARD
Test hypothesis that exposure is associated with disorder (epidemiological/laboratory studies)
Define criteria for EARD
Refine criteria for given EARD

Briton LA et.al., AA Epidemiol 2001;12 321-326
Mortality among augmentation mammoplasty patients .[SIRG(01)40]
This was a retrospective cohort study covering patients from 18 plastic surgery practices in six geographic areas. All women who had bilateral augmentation mammoplasty between 1960 and 1988 were included, but women with breast cancer were excluded. A total of 13,488 subjects were matched with 3936 women who had other forms of plastic surgery. Information about alive/dead/cause of death was obtained for 80% of both groups. Mortality rates and SMR were calculated. The average age of the implant group was 34.0 years compared to 40.9 years for the control. There was a lower overall mortality for both implant and control groups compared to the general population (US rates). Implant patients had excess risks of death for brain cancer and suicide, compared with the general population, and also lung cancer when compared to the control group. The excess mortality from respiratory tract cancers may relate to higher rate of cigarette smoking. A major finding though is that women having plastic surgery fare substantially better than the general population.

Briton LA et.al., AA Epidemiol 2001;11 248-256
Cancer Risk at sites other than breast following augmentation mammoplasty. [SIRG(001)20]
Objective: The objective of this study was to determine whether there was any excess cancer incidence and cancer mortality in women receiving a breast implant. Subsidiary objectives included an investigation in possible (cancer) site-specific increases, and whether any increase is related to type of implant, duration since implant and possible lifetyle and other differences between implanted and non-implanted women.

Design, Subjects and Methods: A retrospective cohort study was used, with cohorts of women receiving implants before 1989 identified and their current status determined . Tracing was successful for 10778 of the 13488 women identified and the following information obtained: (i) their vital status (dead/alive), (ii) cause of death in those deceased, and (iii) personal history of cancer obtained by questionnaire or interview in 71% of those traced. Medical records were obtained where possible to verify diagnosis in those either reporting a cancer of with cancer on the death certificate. These data were compared with (i) both local cancer incidence rates (approximating to the area of residence of the clinics) and national cancer mortality data, these comparisons were adjusted for age, race and calendar year and (ii) a specially recruited cohort of patients from the same clinics receiving other plastic procedures during the same period. These latter controls were also surveyed to obtain data on potential confounders and might therefore be considered a better match. Only a 30% matching was planned for and 3214 of 3936 eligible controls were traced, of whom 71% provided survey data on potential confounders. An identical follow-up procedure was adopted. Statistical comparisons were undertaken using standardised incidence rate ratios, comparing the incidnce rates with local expected rates; 95% confidence intervals are provided for some of these. Mortality comparisons were analysed using SMRs. The comparison with the control rates was undertaken using a Poisson Regression adjusting for potential confounders.

Results:
1. Incidence comparison with local rates
These data are shown in Table 3 of the paper as 'external analyses'. For all sites there was a 20% increased risk (SIR 1.21, 95% CI 1.1-1.4). The most important numerically was in the Cervix with a three fold increase and in the brain with a twofold increase, both of which were 'statistically significant'. None of the major sites showed any decreased risk.

2. Mortality comparison with national rates
These data are shown in Table 5 of the paper. Overall there was a reduction in cancer mortality: SMR 0.77 (95% CI 0.6-0.9) with only brain showing an increase, SMR 2.52. Cancer mortality from breast cancer was significantly reduced in this cohort.

3. Incidence comparison with plastic surgery controls
These data are shown in Table 3 of the paper as 'internal analyses' The rate of cancer, adjusting for age, year and race, was identical in both groups RR1.0, 95% CI 0.8-1.2. Within this overall total there was a non-significant increase in risk of cervical cancer of 1.8 and an increased risk of brain cancer of 2.8. Many sites showed a reduction in the implanted, compared to the non-implanted subjects. The important positive finding was of a 2.2 fold, statistically significant, increased risk of lung cancer in the implanted patients, which persited (data not provided) after adjustment for self reported smoking.

4. Mortality comparison with plastic surgery controls
These data are shown in Table 5, final columns, and shows a similar increased risk of lung cancer, possibly also of brain cancer, but no overall mortality risk

5. Other findings
These was no relation to type of implant with any cancer. Medical record follow-up suggested that the brain cancers were primary, mainly glioblastoma multiforme.

Limitations of Study: The study was strong methodologically as might be expected from the well respected authorship. The external comparisons are troublesome as national mortalty rates are not likley to be appropriate. Similarly the Atlanta cancer incidence data, might not be an appropriate comparison group for these implanted women. Indeed the authors conclude in relation to the cervical cancer women: "Thus our findings suggest that previously observed cervical cervical cancer excesses among implant patients are probably more attributable to reproductive and lifestyle factors common to women undergoing plastic surgery than to the effect of silicone exposure" Thus it may not be possible to adjust for all possible confounders. The non-tracing/non-participation rates were higher than would have been the case in a UK study, given the better national mortality and cancer systems in the UK. However, this is unlikely to have introduced bias between the internal controls. It is possible that tracing deaths may be easier than those still alive and hence they may have over-represented mortality overall and hence cancer mortality in the groups. The authors also comment on the difficulties in obtaining tissue diagnoses for the malignancies, again this would not have biassed the results in any particular direction. The authors tested several simultaneous hypotheses and did not adjust their p values for such testing: Over 20 cancer sites were compared for incidence and mortality with two control groups, ie 80 comparisons, thus by chance alone 4 would have emerged as positive with a p value of <0.05.

Implications of the Study: In general there is no overall suggestion that implants are assocaited with an increased risk of cancer incidence and mortality across all sites. There was a consistent increaseed risk of brain cancer, compared to both control groups, which increases the robustness of that result. Overall, the incidence of these cancers of the brain was 1/8000 person years of follow up, about twice that expected by chance. There is no biological explanation for the finding. There was also evidence of an increase in respiratory and other smoking-related cancers, which persited after adjustment for smoking but might still be explained by smoking, given the weak quality of their smoking data. Finally, the absence of an increase of non-solid tumours such as lymphomas is reassuring.

Summary: This large and carefully conducted study is broadly reassuring on cancer risk following implants. An increased risk at some sites, as a consequence of the implant per se, remains, particulalry in rare brain tumours.

Fryzek JP et.al., Plastic & Reconstructive Surgery 2001 vol 107 214-221 Local complications and subsequent symptom reporting among women with cosmetic breast implants [SIRG(001)21]
Objective: The objective of this study was to estimate the burden and predictors of local complications following breast implant surgery and whether psychological/physical symptoms are linked to these.

Design, Subjects and Methods: Retrospective population based cohort study of cosmetic breast implant patients in Sweden. A random sample was taken of 2500 women who had implants between 1969-1996 with a comparison group (random sample of 3500) who had undergone breast reduction surgery (frequency matched for age and calendar year of surgery). Women were excluded due to previous mastectomy or breast cancer thus 5% of the implants and 1% of the controls were excluded. Patients were sent a questionnaire about 28 specific symptoms and demographic/lifestyle information. Non-responders (after 1 mail reminder) were telephoned and an interview conducted (if consenting) over phone. Response rates were 65% of implant and 72% of breast reduction patients. Further patients were excluded due to CTDs, cancer or unknown height and weight. Thus 1369 women with breast implants and 2211controls were studied. Local complication rates following surgery were extracted from the medical records. Relative risks were calculated on the association between a set of the self-reported symptoms and each complication category (none, any, capsulotomy, implant change and leakage).

Results:
Characteristics of patients with local complications:
Patients with any local complications (capsulotomy, change or leakage) were more likely to be over 35yr at time of operation (RR=1.6), have smoked cigarettes (RR=1.3), had saline implants (RR=2.2), had bleeding (RR=2.2) or infection (RR=6.8) during initial hospital stay for procedure and taken antibiotics after the procedure (RR=1.5).
There were also differences in the likelihood of specific complications depending on these and other surgery-related variables.

Relative Risk estimates for 28 symptoms
Given 28 symptoms and three complications there were 112 associations reported (including "all complications group"). However, a lack of confidence intervals or adjustment for multiple comparisons renders most of the data interpretable.
Compared to the implanted women without local complications, of the 128 associations reported, only three had a RR of over 1.5, only one of which was p<0.05. there were however more positive (rr>1) than negative (RR<1) associations. there were more positive rrs when compared to the breast reduction group. rrs were increased for 12 symptoms in the group with more than one of the three local complications compared to the group with one complication. none was significantly different. RRs were increased for 22 symptoms in the group with >1 capsulotomy or implant change compared to the group with 1 capsulotomy or implant change. None were significantly different.

Limitations of Study:
This was a large well-defined study population with medical record validation of local complications. The failure to take account of multiple comparison is a significant weakness. The final participation rates were low.

Summary:
An important proportion of implant patients had a documented local complication after their surgery. There may be an association between local complications and self-reported symptoms but these are weak.

Fryzek JP et.al., Plastic & Reconstructive Surgery 2001 vol 107 206-213 Self-reported symptoms among women after cosmetic breast implant and breast reduction surgery [SIRG(001)22]
Objective: The objective of this study was to evaluate the possibility that an individual symptom or a collection of symptoms is related to silicone implants after implant surgery. This report is based on the same material discussed in SIRG(001)-21.

Design, Subjects and Methods: See SIRG(001)-21. The analysis for this study compared symptom frequency in both groups.

Results: Women with breast implants had higher mean age (44.3yr) and a significantly lower body mass index (BMI) (21.0) compared to the breast reduction group (mean age=42.2yr and BMI=24.5). The implant group was also more likely to have smoked, to have been pregnant and to have been treated in a university hospital compared to the reduction group. As with SIRG(001)-21, no consideration was given to multiple testing.

Breast implant status and symptom occurrence
Relative risks (RR) generally exceeded 1.0 for the implant group indicating that they generally reported symptoms more often than the reduction group. The RR was over 1.5 only for 5 symptoms which in all cases was significant (p<0.05). this group also reported more symptoms (average=3.0, compared to 2.4 in the reduction group). 75% of the women in both groups reported up to 5 symptoms.
for those with implants that reported symptoms, the mean onset from operation to symptom was 6.3yr compared to 6.1yr in the reduction group.

Association of implants and symptoms by time since operation
No obvious pattern of increasing or decreasing RR according to time was noted.

Association of implant size with symptom reporting
Restricted to the implant group - trends for muscle pain and weakness, and depression/low-spirits were noted but they were not increased relative to increases in implant size.

Association of implant type with symptom reporting
Implant type information was available for 83% of women and small saline implants were used as the reference group. Statistically significant risks were found for:

Large saline implants: excess dry mouth (RR=5.1, 95% CI 1.1-24.4), other skin abnormalities (RR=3.9, 95% CI 1.5-10.2 )
Small silicone implants: excess dry eyes (RR=2.9, 95% CI 1.1-7.9)
Large silicone implants: persistent/recurrent shoulder ache (RR=1.6, 95% CI 1.0-2.6).

Limitations of Study: Bias was reduced by matching implant and reduction patients by calendar year of surgery thus placing similar memory demands on the two groups.

Implications of the Study:

• Women with implants report symptoms more often than women who have undergone breast reduction surgery. They propose 4 explanations:
• • Implant patients may have the sort of personality that makes them more likely to notice and report symptoms
  • Recall or reporting bias e.g. from media reporting may cause an excess of symptom reporting in this group
  • Selection bias - women with the symptoms may be more likely to reply to the study questionnaire
  • The control group may have had a procedure that means some of the symptoms may be reduced due to the surgery.
• As such a broad range of symptoms were reported it does not seem to tie-in with a distinct exposure (silicone).
• Symptom reporting had no definite pattern with regards to time since operation, nor was it increased over time or higher close to surgery - average length was 6yr from surgery to report of symptom for both implant and reduction groups.
• Mass/Surface area of implant was not related to symptom occurrence and no apparent difference between reporting profiles of women with large saline implants than those with other silicone implants.
• This is the largest study (to date) to look at symptom reporting in breast implant patients. The findings support recent work in Canada of cosmetic surgery patients and also a small study in the Netherlands which found breast implants patients more than twice as likely as other cosmetic surgery patients to report symptoms.
• If an atypical syndrome related to silicone does exist it is yet to be found by either symptom reporting or immunological parameters

Krueger EA et.al., Int.J.Radiation Oncology Biol.Phys 2001 vol49 713-721 Complications and patient satisfaction following expander/implant breast reconstruction with and without radiotheraphy. [SIRG(001)23]
81 patients were received 19 having had radiotherapy and 62 never having had radiotherapy. Those having had radiotherapy had a complication rate of 68% and a reconstruction failure rate of 37% compared to those without radiotherapy who had a complication rate of 31% and a reconstruction failure rate of 8%. Patients were generally satisfied with the reconstruction aesthetically irrespective of whether they had radiotherapy or not.
The increased failure and complication rate following radiotherapy has been well documented in the past. The authors point out that the numbers are the small for statistical significance but as a pilot study they feel it points the way.

Benadiba L., Ann Chir Plast Esthet 2000 Feb 45:1 31-40
Survivorship of breast implants used in breast reconstruction. 949 implants [SIRG(001)31]
This French language paper gives a global median life span for implants of 127 months (10¾ years). The authors observed that gel implants last longer than saline though this may be confused by the fact that gel implants can rupture asymptomatically and so not be detected. They state that under-filled saline implants don't last as long as well-filled ones (108 months compared to 127 months). They also show a graph to show variable life span for saline implants depending on the manufacturer.

Kamel et al J Biomed Mater Res (Applied Biomarkers) 91; 58: 88-96. The peri-implant breast capsule: An immunophenotypic study of capsules taken at explantation surgery, [SIRG(001)34]
Capsules from 63 patients, removed between 1992 and 1996, were examined for the presence of T and B lymphocytes and macrophages, using immunohistochemistry, as well as for collagen IV, S.M.A. and CD34. The number of positive cells in 10 consecutive high power fields was assessed, and related to clinical data.

Implants had been in place 4 - 25 years with a mean of 14 years, and were removed because of suspected failure, induration, pain or fear of complications.
There was an almost complete absence of B cells and activated T cells, with T cells and macrophages predominating. There was no correlation between duration of implant and either the extent of infiltration or clinical features, nor was there any specific pattern of response to suggest an immune or sensitivity process.

Smalley DL et al., Pathology 1998 66:302-305
Monocyte-Dependent Stimulation of Human T cells by Silicon Dioxide [SIRG(001)35]
Mononuclear cells from silicone breast implant patients were cultured in replicates either with Concanvalin A (Con A) or silicone dioxide, and the T cell proliferative responses determined by the uptake of tritiated thymidine. Experiments were performed to determine the effects of monoclonal antibodies to HLA-DR, HLA-DQ, HLA-DP, and to B7-1 (CD 80) on the proliferative responses. Monoclonal antibodies to DQ, DP, and CD 80 inhibited the proliferative responses, as did one of two anti-DR antibodies tested. The authors concluded that the results showed that the T cell proliferative response to silicon dioxide is monocyte-dependent, unlike the response to superantigens.

Comment: As mentioned in the original detailed scientific report from the Independent Review Group, published on the web (see section 3 of Immune responses to silicone gel breast implants in www.silicone-review.gov.uk/immune), there is a need for independent laboratories to confirm or refute Dr Smalley's claim that silicone dioxide induces a memory T cell proliferative response in lymphocytes taken from women with silicone breast implants. This paper re-emphasises the need for such further research.

Dubler N N, Schissel A. JAMWA 2000 55:5 255-6 Women, Breasts and the failure of informed consent. [SIRG(001)37]
This article was written at a time when patient autonomy, the protection of patients from unscrupulous doctors, and the issue of consent to medical treatment and research were issues of concern at the highest level in the United States of America. President Clinton just had expressed great concern about what he perceived to be insufficient protection of human subjects in research, and Secretary Shalala had announced the transfer and expansion of human research protection to the newly established NIH Office for Human Research Protections. That Office is responsible for developing and monitoring regulations for the protection of human subjects and dealing with training and certification of researchers.

The article, which focuses in particular on the recent review of saline breast implants in the United States, makes the valid point that if there is a conflict of interest involving the tension between commerce and medicine, the North American doctrine of informed consent (1) fails to provide adequate protection for patients. Accordingly, surgeons who have a financial interest in selling successfully recommending breast implants to women, might be deterred from informing their patients of the possible risks and side effects of breast implant surgery. The authors point out that statements of possible risks, complications and side effects must be stated at the same time as promises of successful outcomes. The article contains examples of advertisements for breast implants designed to enhance the commercial market for cosmetic breast implant surgery.

The doctrine of informed consent developed in North America, but not accepted in its entirety in the UK, is clearly explained by the authors, who argue that, while the doctrine does support patient autonomy as against medical paternalism, it does not adequately deal with the problem of combating sophisticated advertising forces. Their views are summed up with the words:

"With stimulated demand, informed consent for breast augmentation is too fragile a reed to withstand the storm of commerce".

The solution suggested by the authors is that advertisers of medical devices be mandated to disclose in their advertisements statements of side-effects, complications and other adverse factors. They recommend the use of the model that was due to be introduced in a Bill in the USA in June 2001 to encourage compliance with FDA requirements for prescription medicines by denying tax benefits for unbalanced advertisements that placed undue emphasis on their benefits without also explaining the relevant risks, or failing in some other way to comply with FDA requirements.

The UK position
The article refers only to the legal position in the United States. UK law relating to consent to medical treatment differs in some respects from US law. In particular, there is no legal requirement in the UK for a clinician to provide full information to patients who are about to undergo medical treatment, about any risks or side effects of that treatment. However, UK Government sources state that doctors are advised to provide information to patients about these matters (2). The law as reviewed by the House of Lords in the leading case in this matter is not expressed clearly (3), and their Lordships were not in complete agreement on the legal position. Subsequently, there have been indications that the UK courts are developing the law in such as way as to bring it closer to the US model. However, the law is still in the process of development and there are conflicting decisions (4) concerning consent to treatment in recent years. Despite cumulative trends towards greater recognition of patient autonomy by the UK judiciary, it may be some time before a definitive statement of the law on consent is made by the House of Lords.

Despite the differences in the approach of the law in the UK and the USA, it is encouraging that the recommendations of the Independent Review Group in the UK, and the recently published patient information leaflet based on the those recommendations, answer many of the concerns expressed by the authors of the article. Women in the UK who are considering having silicone gel breast implants are advised by the leaflet to ask specific questions of the surgeons about the risks, potential complications and side effects of surgery, and information is provided in the leaflet itself to enable women to reach informed decisions. Concerns were expressed by the Review Group in its report in 1998 about the persuasive power of advertising propaganda and the need to introduce controls over clinics in the independent health sector was emphasised. Since that report was published, the Government has enacted legislation to monitor and regulate the independent sector and to bring it in line with NHS provision (5). It is hoped that all these measures will combine to contribute to improving the protection of patients who are considering, and who undergo, breast implant surgery.

Kaufman HH, Science and Justice 2001, 41(1) 7-20
The expert witness. Neither Frye nor Daubert solved the problem: what can be done? [SIRG(001)38]
This article deals with the role of US judges as "gate-keepers" of scientific evidence in the US Courts. The author takes the view that because judges do not understand fully the scientific issues, they have problems with this role, and contends that further measures are required to ensure that judges do not reject some scientific evidence unjustifiably. The improvements recommended by the author include more efficient procedural systems using neutral experts and a mechanism for holding experts accountable for any improprieties they are found to have committed.

The paper refers to instances of medical malpractice litigation in America and the UK, and to mass torts, covering the silicone gel breast implant controversy. The author points to the cases of Frye (6) and Daubert (7) , which created standards by which to evaluate the reliability of scientific testimony, and improved the value of scientific evidence by giving judges the opportunity to assess it before it is adduced. However, he emphasises that, despite these factors, experts can still be biased, and there is still some potential for "junk science" to find its way into court.

The author explores the reasons why scientific evidence may be flawed, concluding that some erroneous information is given purposely, while other scientific information is based on lack of knowledge by experts, many of whom, he alleges, receive handsome fees for very little work.

The author offers some solutions aimed at improving the quality of expert scientific testimony, and points out that some of the problems have already been solved by more sophisticated techniques of medical research and greater knowledge of scientific matters. Concerns expressed by the scientific community and by lawyers have resulted in greater awareness of the problems that can arise. He suggests that attitudes to scientific testimony in the USA need to change so that the legal system becomes far more selective in permitting experts to testify, and testifying becomes a duty and an honour accorded only to the best possible expert witnesses who are fully trained in legal principles and processes. He recommends that experts be required to explain, during the pre-trial process, how and why their testimony differs from that of opposing experts. He suggests that fees payable to experts be limited so that they are not biased by financial considerations, and more innovatively, he proposes that judges should be trained in scientific matters, or even that only scientifically trained and dually qualified judges should hear cases concerning complex scientific issues. He also recommends that opposing experts on a particular scientific matter be required to testify alternately, in order to allow their evidence to be evaluated contemporaneously, and that there should be more court appointed experts. He particularly recommends the approach taken in the course of the silicone gel breast implant litigation, when Judge Pointer successfully used a scientific panel to assess in advance the quality of the evidence to be presented to the jury.

The UK Position
The position in civil cases in the UK is very different from that in America. In the UK, civil claims for damages are not heard by juries as they are in America, and judges are considered better able than juries to assess expert evidence. The legal establishment in the UK has been well aware for some time of the potential problems of using experts, and when Lord Woolf conducted his radical review of the Civil Justice System in the 1990s (8), he concluded that the use of expert witnesses was one of the major generators of costs in civil proceedings. The result of that review has been a controversial new system for the conduct of civil proceedings, one result of which is that UK judges now have much stronger control over the conduct of case, the use of expert witnesses and the evidence they present to the court (9).

The Civil Procedure Rules 1999 place the judge in complete control from the pre-trial stage, and this includes control over the fees payable to experts. In many cases expert testimony is given only in writing, and the number of experts acting in cases has been reduced considerably. If the parties are unable to agree on an expert the court can appoint one from an approved list. The Rules contain statements of the duties of expert witnesses and experts are penalised if they fail to comply with their duties. All experts have a duty to the court and not primarily to the party for whom they are testifying (10), and their written testimony must contain a sworn "statement of truth", as follows:

"I believe that the facts I have stated in this report are true and that the opinions I have expressed are correct".

Expert scientific witnesses are required to undergo training and to belong to recognised associations. The Lord Chancellor's Department in the UK has recently carried out a preliminary review (11) of the operation of the Civil Procedure Rules, and this concludes that the Rules relating to expert witnesses are working satisfactorily, and that there is now a less adversarial approach as between the parties to litigation. Further research is on-going.

The Human Rights Act 1998 incorporates the provisions of the European Convention on Human Rights into UK law. Article 6 of the Convention states that everyone is entitled to a fair hearing in the determination of their civil rights and obligations. This provides additional safeguards, as improper use of experts would infringe Article 6.

Conclusion
While there continue to be concerns about the use of expert witnesses and the validity of scientific evidence in the United States of America, the position is very different in the UK, where measures are in place to ensure that the courts have control over these

Footnotes
1. See Schloendorff v Society of the NY Hospital, 211 NY 125, 105 NE 92 (1941) and the Patient Self-Determination Act, Omnibus Budget Reconciliation Act 1990, Pub L No 101-508, 4206, 4751 Stat. 1388, 1388-155, cited by the authors on page 256, op cit

2. See "Consent: Reference Guide To for Examination or Treatment", Department of Health 2001

3. Sidaway v Governors of Bethlem Royal Hospital (1985) 1 BMLR 132

4. c.f. Pearce v United Bristol Healthcare Trust (1998) 48 BMLR 118, and Burke v Leeds Health Authority, Unreported, Court of Appeal 29th January 2001

5. The Care Standards Act 2000 and the Health and Social care Act 2001

6. Frye v United States, 54 App. D.C. 46, 293 F. 1013, 1923

7. Daubert v Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579-601, 1993

8. Lord Woolf produced an Interim Report and a Final Report in 1996 and 1998 respectively

9. See Part 35 CPR and Practice Directions for detailed information on the use of expert witnesses

10. CPR 35 (4)

11. Emerging Findings: An Early Evaluation of the Civil Justice Reforms, Lord Chancellor's Department, March 2001