EXP BREAST NEOPLASMS

ALL PROPHYLAC: OR
ALL PROPHYLAXIS OR
ALL HYPERTROPHY OR
ALL SUBCUTANEOUS

EXP MASTECTOMY OR
EXP MAMMAPLASTY OR
ALL MAMMOPLAST: OR
ALL MAMMAPLAST: OR
LYMPH ADJ NODE ADJ DISSECTION

EXP BREAST NEOPLASMS

PROPHYLACTIC ADJ SURGERY OR
PROPHYLACTIC ADJ SURGERIES OR
PREVENTIVE ADJ RESECTION OR
PREVENTIVE ADJ MASTECTOMY OR
PROPHYLACTIC ADJ TREATMENT#

PROPHYLACTIC (TF)

MASTECTOMY (TF)

EXP BREAST NEOPLASMS

ALL RISK

SU OR EXP MASTECTOMY

GENES, BRCA1 OR
GE OR
GENETIC SCREENING OR
HYPERTROPHY/SU

BREAST(W)CANCER
BREAST(W)CARCINOMA

PROPHYLAC?
PROPHYLAXIS
HYPERTROPHY
SUBCUTANEOUS

MASTECTOMY
MAMMOPLAST?
MAMMAPLAST?
LYMPH(W)NODE(W)DISSECTION

PARTIAL/TI,DE
RECURRENCE/TI,DE
COUNSELING/TI,DE
BILATERAL/TI,DECANCER(W)SURVIVAL
CANCER(W)MORTALITY

BREAST(W)CANCER
BREAST(W)CARCINOMA

PROPHYLACTIC(W)SURGERY
PROPHYLACTIC(W)SURGERIES
PREVENTIVE(W)RESECTION
PREVENTIVE(W)MASTECTOMY
PROPHYLACTIC(W)TREATMENT?

PARTIAL/TI,DE
RECURRENCE/TI,DE
COUNSELING/TI,DE
BILATERAL/TI,DE
CANCER(W)SURVIVAL
CANCER(W)MORTALITY

BREAST(W)CANCER
BREAST(W)CARCINOMA

RISK

SURGERY
MASTECTOMY

GENES
GENETIC
HYPERTROPHY

PARTIAL/TI,DE
RECURRENCE/TI,DE
COUNSELING/TI,DE
BILATERAL/TI,DE
CANCER(W)SURVIVAL
CANCER(W)MORTALITY

BREAST CANCER/DE
BREAST CARCINOMA/DE
BREAST NEOPLAST?/DE

PROPHYLAC?/TI,DE
PROPHYLAXIS/TI,DE
HYPERTROPHY/TI,DE
SUBCUTANEOUS/TI,DE

MASTECTOMY/TI,DE
MAMMOPLAST?/TI,DE
MAMMAPLAST?/TI,DE
LYMPH(W)NODE(W)DISSECTION

BREAST CANCER/DE
BREAST CARCINOMA/DE
BREAST NEOPLAST?/DE

PROPHYLACTIC(W)SURGERY
PROPHYLACTIC(W)SURGERIES
PREVENTIVE(W)RESECTION
PREVENTIVE(W)MASTECTOMY
PROPHYLACTIC(W)TREATMENT

BREAST CANCER/DE
BREAST CARCINOMA/DEB
REAST NEOPLAST?/DE

RISK/TI,DE

SURGERY/TI,DE
MASTECTOMY/TI,DE

GENES/TI,DE
GENETIC/TI,DE
HYPERTROPHY/TI,DE

PROPHYLACTIC/(w) MASTECTOMY

Study

Study Design

Selection bias

Performance bias

Detection bias

Attrition bias

Was selection bias (or preferential selection in case series) satisfactorily minimized?

Selection bias was satisfactorily minimized if the major known risk/protective factors (confounders/co‐interventions) were controlled for to ensure comparability between groups. For case series preferential selection was minimized if there was evidence of a consecutive sample, or a clearly defined patient population to minimize the chance that clinicians selected only patients with favorable outcomes or that patients with better outcomes volunteered (healthy volunteer bias).

Performance bias satisfactorily minimized?

Performance bias was satisfactorily minimized if prophylactic mastectomy was confirmed in an objective way (i.e., medical or surgical records) and not based exclusively on self‐report.

Was detection bias satisfactorily minimized?

Detection bias was satisfactorily minimized if the outcome was assessed in a valid way (i.e., validated pre/post instruments for psychosocial measures, medical records for incidence, medical/death records for vital status), in the same way for both groups, and by a blinded outcomes assessor.

Was attrition bias satisfactorily minimized?

Attrition bias was satisfactorily minimized if there was there a low dropout rate and/or if dropouts/withdrawals were sufficiently accounted for so that the reviewer was convinced that differential reasons for dropping out did not occur or were accounted for.

Mulvihill 1982

Retrospective cohort

No. People are included in this study for 'risk factors' that would not be considered risk factors for breast cancer by today's standards. Major confounders are not controlled for.

Yes

Don't know (not enough details provided)

Yes

Stefanek 1995

Prospective cohort

Yes

Yes

No. There is nothing in this article stating that a validated patient satisfaction instrument has been used. Biases can result from not using a validated patient satisfaction questionnaire that skews towards favorable outcomes (Rubin 1991).

Yes

Borgen 1998

Convenience sample

No. Because the participants responded to advertisements, those who responded could be different in some important way than those who did not respond.

Yes

No. There is possible recall bias from collecting all psychological data postoperatively.

Yes

Hartmann 1999

Retrospective cohort

No. Study does not control for other important preventive measures besides PM that may have been used by this population including oophorectomy and chemoprevention.

Yes

Yes

Yes

Hatcher 2001

Prospective cohort

No. 5% have had genetic test in the decliners vs. 29% in acceptors. We do not know if there are different baseline risks.

Yes

Yes

Yes

Hartmann 2001

Case series

Yes

Yes

Yes

Yes

Frost 2001

Case series

Yes

Yes

No. There is possible recall bias from collecting all psychological data postoperatively.

Yes

Meijers‐Heijboer 2001

Prospective cohort

No. While the authors controlled for some factors such as age and oophorectomy status, adjustment of other important factors is not reported.

Yes

Yes

Yes

Josephson 2000

Case series

Yes

Yes

No. These interviews took place at least 1 year after the procedure (1 took place after more than 3 years, 3 took place after 2 years, and 9 after more than 1 year). Yet, women were asked to recall their satisfaction with the information provided by the surgical/reconstructive team and preoperative genetic counseling. By the authors' own admission on page 354 "there is a risk that the women may not properly recall how and what they experienced at the time."

Yes

Lloyd 2000

Case series

Yes

Yes

No. In qualitative studies, the 'validity' of the domains is, in part, determined by interviewing enough patients and for enough time that informational redundancy is reached. This is not mentioned.

Yes

Hopwood 2000

Case series

Yes

Yes

No. This is entirely retrospective data. Baseline measures have not been collected. There is possible recall bias from collecting all psychological data postoperatively.

No.

Zion 2000

Case series

Yes

Yes

Yes

Yes

Leis 1981

Case series

No. People are included in this cohort who had benign diseases that are not considered risk factors for breast cancer by today's standards.

Yes

No. Valid disease‐free survival estimates depend on all patients getting assessed for disease at regular, fixed intervals. It is not mentioned whether this occurred.

No. There are 69/127 patients who have not been accounted for.

Babiera 1997

Retrospective cohort

No. There is a chance of selection bias. Those electing CPM may differ in some important way from the control group and important factors have not been adjusted for in the analysis. For example, there is a median age difference of 7 years between the two groups. Also, it is stated that the decision to have CPM was influenced by histology and family history: " We do not know how histology and family history were different between the two groups."

Yes

No. Incidence in contralateral PM group is given as occult incidence at time of surgery with no subsequent follow‐up data provided, whereas incidence in control group is given up to 103 months. This appears to be an apples and oranges comparison. Also, because valid disease‐free survival estimates depend on patients in both groups getting assessed at the same fixed intervals, we do not know the validity of these data as intervals are not mentioned.

Yes

McDonnell 2001

Case series

No. The analysis does not sufficiently control for confounding factors, i.e., histology or stage of primary tumor.

Yes

Yes

Yes

Montgomery 1999

Convenience sample

No. Because the participants responded to advertisements, respondents may be different in some important way than the nonrespondents.

Yes

No. There is a possibility of recall bias by asking all quality of life questions after the surgery only.

No 50 women of 346 did not respond to questionnaire

Peralta 2000

Retrospective cohort

No. This study does not adjust for all major confounders.

Yes

No. Valid disease‐free survival estimates depend on patients in both groups getting assessed at the same fixed intervals. Therefore, we do not know the validity of these data as intervals are not mentioned.

Yes

Lee 1995

Retrospective cohort

No. The study only adjusts for age, not the other major confounders. Also, the treatment group includes those undergoing contralateral PM as well as those having biopsies. It is unclear how including those with only biopsies may have biased the results.

Yes

No. PM group is combined with those receiving biopsies; therefore, the risk in the PM group is not ascertainable.

Yes

Horton 1979

Case series

No. People were included in this cohort who had benign diseases that are not considered risk factors for breast cancer .

Yes

Yes

No

Pennisi 1987

Case series

No. People are included in this cohort who had benign breast diseases that are not considered risk factors for breast cancer by today's standards.

Yes

Don't know

No. The 30% loss to follow‐up increases risk of attrition bias.

Evans 1999

Case series

No

Yes

Don't know. It is uncertain how the estimation of risk is affected by use of the Claus model.

Yes

Gabriel 1997

Case series

Yes

Yes

Yes

Yes

Study

Outcome

Length of followup

Attrition

Study Details

Meijers‐Heijboer 2001

TREATMENT GROUP:
BPM group:…………..0/76

COMPARISON GROUP:
Surveillance group: 1/63

RR=0.28 (95% CI 0.01, 6.68) p=0.43

Mean followup of 3.0 +/‐1.5 years.

None

See Table 5 for study population details.

Hartmann 1999

WOMEN AT HIGH RISK

TREATMENT GROUP:
BPM: ……….2/214 deaths

COMPARISON GROUP: …. 90/403 deaths
Using three different methods to calculate incidence taking into account ascertainment bias, the risk of death was reduced by 81‐94%.

Most conservative estimate for high risk:% reduction=80.9% (95% CI 31.4%, 97.7%)

Moderate risk:
BPM:……….0 of 425
Predicted incidence of death: 10.4 of 214

% reduction=100% (95% CI 70%, 100%).

The median length of followup was 14 years.

None

See Table 5 for study population details and definitions of "high risk" and "moderate risk.".

Pennisi 1987

3 of the 1500 BPM/CPM patients died from breast cancer.

No comparison group

70% followed for 9 years.

30% were lost to followup

1500 patients from 165 plastic surgeons who had subcutaneous PM and were registered with the Subcutaneous Mastectomy Data Evaluation Center.
78 (5.2%) subjects had obscure carcinoma and 51 (3.4%) had LCIS at the time of surgery and were included in the study.

Among the 139 patients who had CPM, 4 (3%) had breast cancer and 5 (3.6%) had LCIS and were included in the study.

300 (20%) had a first‐degree relative with breast cancer and 21% had a history of second‐degree maternal or paternal relatives with a history of breast cancer.

Skin necrosis occurred in 5% of the patients.

Study

Incidence

Length of followup

Attrition

Study Details

Meijers‐Heijboer 2001

Bilateral Prophylactic Mastectomy (BPM)

TREAT MENT GROUP:
BPM…………………0/73

COMPARISON GROUP:
Surveillance arm…...8/63

BPM significantly (P=0.003) decreased incidence of breast cancer at 3 years' follow‐up. Hazard ratio=0 (95% C.I. 0.0 , 0.36).

Mean followup of 3.0 +/‐1.5 years

None

Using the surveillance group, the authors estimate the five‐year risk of breast cancer was 24 +/‐ 9%.

The ratio of observed occurrences to expected occurrences in the surveillance group was 1.2 (8 vs. 6.7).

Significantly more women in the BPM arm than in the surveillance arm also had an oophorectomy (44 vs. 27 [58% vs. 38%]).

The 139 women were from 70 families.
None of the affected women were from the same family.

MRI detected 6 of the 6 cancers screened.
Mammography detected 2 of the 8 cancers screened.

Hartmann 1999

Bilateral Prophylactic Mastectomy (BPM)

Moderate‐Risk Women
TREATMENT GROUP:
BPM …….4/425
COMPARISON GROUP:
Statistically simulated using the Gail model......37.4/425.
Risk reduction= 89.5%.

High risk women
TREATMENT GROUP:
BPM …………. 3/214

3 COMPARISON GROUPS
All simulated from probands' sisters breast cancer rates:
1. All breast cancer from age 18 to follwup............ 52.9/214
2. All breast cancer from age 18 to followup corrected for ascertainment bias..............................30/214
3. Only breast cancers that occurred in sisters after probands' diagnosis.........................................37.4/214
Using three different methods to calculate incidence taking into account ascertainment bias, the expected incidents among the 214 high‐risk probands ranged from 30.0 ‐ 52.9/214

Most conservative estimate was %difference=90.0 (95% CI 70.8, 97.9)

Median length of followup was 14 years

None

To be classified as high risk, women had to meet one of the following criteria:2 or more first‐degree relatives with breast cancer;1 first‐degree relative and 2 or more second‐ or third‐degree relatives with breast cancer;1 first‐degree relative with breast cancer before the age of 45 and one other relative with breast cancer;2 second‐degree or third‐degree relatives with breast cancer and 1 or more with ovarian cancer;1 second or third‐degree relative with breast cancer and 2 or more with ovarian cancer;3 or more second or third‐degree relatives with breast cancer;1 first‐degree relative with bilateral breast cancer.

2 women in the high‐risk group developed ovarian cancer.

All 7 who developed breast cancer had subcutaneous mastectomies. But there was no significant difference in outcome between the group with subcutaneous mastectomies compared to those who had total mastectomies.

Median time to development of breast cancer was 6 years.

At the time of the study, tissue was available for pathological review for 603 of the women. 2 invasive cancers were identified during the review. 1 of the 2 women had developed breast cancer 3 years after the PM

Mulvihill 1982

Bilateral Prophylactic Mastectomy (BPM)

TREATMENT GROUP:
BPM……… … 0/9

COMPARISON GROUP:
Surveillance ……0/4

1 to 12 years

None

The 2 groups were similar in age, prior attitudes, and knowledge of breast cancer and control measures.

None of the 4 women who chose surveillance had followed minimum of monthly self‐examination and quarterly examination by a physician

Hartmann 2001

Bilateral Prophylactic Mastectomy (BPM)

Patients with BRCA1/2 mutations

TREATMENT GROUP:
BPM……………….0/26

COMPARISON GROUP:
Simulated group #1 using Easton penetrance model.. 9.37/26
Risk reduction =100% (95% C.I. 51.0‐100.0).

Simulated group #2 using Struewing penetrance model= 6.52/26
Risk reduction = 100% (95% C.I. 54.1‐100.0).

13.4 years (range 5.8‐28.5 years).

None

Subjects are a subset of the 214 high‐risk women who were participants in Hartmann's 1999 study. 26 had alterations in BRCA1 or BRCA2.

8 of the original 214 subjects in the cohort had died at time this study began: 2 from breast cancer, 1 from ovarian cancer. The woman with ovarian cancer had deleterious BRCA1 mutation.

Borgen 1998

Bilateral Prophylactic Mastectomy (BPM)

BPM…………….. 3/370

0.2‐51.5 years with a mean of 14.8 years

Not applicable

255 of 370 (69%) reported the discussion to have PM was initiated by their physician, while 108 (29%) initiated the discussion themselves.

Five did not recall who initiated the discussion.

Incidental carcinoma was identified in 14 of the 370 (4%) and they were included in the study.

Horton 1978

Bilateral (BPM) and Contralateral Prophylactic Mastectomy (CPM)

CPM…………….0/11
BPM …………….0/90

Followup 1 month to 10 years, with an average of 3.1 years.

None

Note: This study contained a small group of CPM patients; however, results are not presented separately. Due to the preponderance of BPM, the study is reported with BPM incidence results.

Pennisi 1987

Bilateral (BPM) and Contralateral Prophylactic Mastectomy (CPM) combined

BPM/CPM………....6/1500

70% of the patients were followed for 9 years.

30% were lost to followup

1500 patients from 165 plastic surgeons who had subcutaneous PM and were registered with the Subcutaneous Mastectomy Data Evaluation Center.

78 (5.2%) subjects had obscure carcinoma and 51 (3.4%) had LCIS at the time of surgery and were included in the study.

Among the 139 patients who had CPM, 4 (3%) had breast cancer and 5 (3.6%) had LCIS and were included in the study.

300 (20%) had a first‐degree relative with breast cancer and 21% had a history of second‐degree maternal or paternal relatives with a history of breast cancer.

Note: This study contained a small group of CPM patients; however, results are not presented separately. Due to the preponderance of BPM, the study is reported with BPM incidence results.

Evans 1999

Bilateral (BPM) and Contralateral Prophylactic Mastectomy (CPM) combined

TREATMENT GROUP:
CPM/BPM…………..…0/400 woman years
COMPARISON GROUP:
Statistically simulated using Claus model……4 /400 woman years were expected

Study followup covered 400 women years for an average of 2.2 years.

None

Women were from 10 European cancer centers that offer risk assessment and counseling services to women with a lifetime risk of breast cancer from 25‐80% using the Claus data.

Authors stated that follow‐up for more than 5 years would be necessary to address the issue of risk reduction.

Note: This study contained a small group of CPM patients; however, results are not presented separately. Due to the preponderance of BPM, the study is reported with BPM incidence results.

Study

Outcome

Followup Time

Attrition

Study Details

Gabriel 1997

Physical morbidity defined as complications leading to unanticipated surgical interventions following breast implantation in cosmetic, prophylactic and cancer patients:

Complications involved 274 (18.8%) of the 1454 breasts with implants and 321 (18.8%) of the 1703 implants.

Complications were defined only as events requiring surgical interventionsBy 5 years, the number of implants with complications was nearly 3 times as high in cancer and prophylactic groups as the cosmetic group.

Cancer group ‐ 34.0% of 125 (95% C.I. 27.2%‐41.3%)
Prophylactic group ‐ 30.4% of 92 (95% C.I. 23.1%‐38.4%)
Cosmetic Group ‐ 12.0% of 532 (95% C.I. 9.1%‐15.2%)

The three most frequent problems were:
1. capsular contracture‐ 17.5% of patients
2. implant rupture ‐ 5.7% of patients
3. hematoma ‐ 5.7% of patients

Mean follow‐up was 7.8 years with a range of 0 to 7.8 years (5847 person years.)

For analysis, follow‐up period was 5 years.

None

208 of the 749 (27.8%) underwent 450 additional surgical procedures within 5 years.

91 of 450 (20.2%) of the procedures were anticipated (staged procedures, patient's request for size change or aesthetic improvement) and 359 had clinical indications and were performed in 178 (23.8%) of the women.

Despite number of complications, author cautions that study did not evaluate patients' overall satisfaction with their implants or the effects of these events on patients' overall health status.

Zion 2000

290 of the 591 (49%) had unanticipated re‐operations (UR).

For all 591 women, the average UR per person is 0.96 (std dev 1.32).

Reasons for UR were
:22% ‐ immediate post‐operative complications
46% ‐ implant‐related issues
32% ‐ aesthetic concerns.

Mean follow‐up of 14.2 years.

None

Physical morbidity assessed by review of medical records and patient interviews to assess complications leading to surgical procedures that were not part of the standard breast implantation protocol.

Median time to UR was 1.3 years with 42% occurring within one year of breast reconstruction.

Of 1182 implants originally placed,
432 (37%) were removed and 389 new implants were placed.

Note: Some of these subjects are probably the same as some of the subjects in the Gabriel study.

Stefanek 1995

7 of 14 (50%) reported they had "quite a bit" or "very much" discomfort after surgery.

2/11 patients had silicon implants remove due to tissue rejection and subsequent infections.

Follow‐up on satisfaction survey from 6‐30 months

None

Study

Outcome

Followup

Study Details

Stefanek 1995

Bilateral Prophylactic Mastectomy (BPM)

DECISION SATISFACTION
14/14 (100%) reported satisfaction with their decision to have PM.

12/14 (86%) reported they would recommend PM to other women with the same risk.

COSMETIC SATISFACTION
Of the 11 opted to have reconstruction,
7/11 (64%) reported being "quite a bit" or "very much" satisfied with results.
1/11 (9%) reported "somewhat satisfied"
3/11 (27%) were "dissatisfied," reporting results were worse than expected.

SURGICAL RECOVERY SATISFACTION
11/14 (79%) were "quite a bit" or" very much" satisfied with the time to recover physically.
3/14 were somewhat or a little satisfied.

13/14 (93%) were "quite a bit" or" very satisfied" with the time to recover emotionally.
1/14 (7%) was somewhat satisfied.

13/13 (100%) reported satisfaction with support from husband or partner.

CANCER WORRY
Women in the BPM group reported more worry about developing breast cancer than women in either of the other 2 groups that elected not to have BPM (P=<0.01).

Women in the BPM group tended to overestimate their risk (based on the Gail model) compared to the group 'not interested' in PM (P=<0.05)

FOLLOWUP
Followup on satisfaction survey from 6‐30 months

ATTRITION: None

Participants were recruited from a group of 164 women with at least one first‐degree relative diagnosed with breast cancer, who were seen by a breast surveillance service (BSS) between 1/1988 and 11/1992.

Satisfaction with BPM assessed through postoperative postal questionnaire.Group comparison analysis:

There was little difference in age, family history, or mammography history among the groups. However, 12 of the 14 had had a recent breast cancer event defined as personal biopsy, diagnosis, recurrence or death of a family member, and biopsy of a family member.

Those selecting PM reported doing breast self‐exam monthly or more frequently, which was significantly more often than those not interested in PM (P=<0.05).

Mulvihill 1982

Bilateral Prophylactic Mastectomy (BPM)

COSMETIC SATISFACTION
All were enthusiastic about the operative results, despite complications in two.

CANCER WORRY
Those who chose mastectomy estimated themselves more anxious and at higher risk than the others.

SEXUALITY
None reported change in sexual activity after BPM.

FOLLOWUP
1‐12 years

ATTRITION
None

Of the 9 counseled prospectively, little distinguished the 5 who chose mastectomy from the 4 who chose surveillance.

They were similar in age, prior attitudes, knowledge of breast cancer and of control measures.

None of the 4 women who chose surveillance had followed minimum of monthly self‐examination and quarterly examination by a physician

Borgen 1998

Bilateral Prophylactic Mastectomy (BPM)

DECISION SATISFACTION
Most women were satisfied with BPM.
21/370 (5%) regretted their decision to have PM, with 19 of them among the 255 for whom the discussion about PM was initiated by their physicians.

Of the 21 with regrets,
10/21 (48%) had major regrets and would not undergo BPM again.
7/21 (33%) had minor regrets.
4/21(19%) did not report level of regrets.

19/21 (90%) of women who were unhappy with BPM results did not have preoperative counseling.

COSMETIC SATISFACTION
Of the 331 who responded about cosmetic results,
116/331 (35%) reported excellent results.
163/331 (49%) reported acceptable results.
52/331 (16%) reported unacceptable results.

FOLLOWUP
The follow‐up years since surgery ranged from 0.2‐51.5 years with a mean of 14.8 years

ATTRITION
Not applicable

Quality of Life/Satisfaction assessed by survey regarding satisfaction and regrets with BPM. There is no mention of whether the survey was validated.

336 patients were selected from a group of 817 volunteers who responded to an invitation in the popular press to join the National Prophylactic Mastectomy Registry. 34 patients were recruited from the authors' practice or the NY Metropolitan Breast Cancer Group.

Women with LCIS were excluded.

220 of the 370 (59%) reported having at least 1 first‐degree relative diagnosed with breast cancer.

255 of 370 (69%) reported the discussion to have BPM was initiated by their physician, while 108 (29%) initiated the discussion themselves. Five did not recall who initiated the discussion.

Mean age of patients with regrets was 45 and group overall was younger than those who were satisfied with BPM.

Incidental carcinoma was identified in 14 of the 370 (4%) and they were included in the study.

Hatcher 2001

Bilateral Prophylactic Mastectomy (BPM)

PSYCHOLOGICAL MORBIDITY/ANXIETY
TREATMENT GROUP:
In the 79 women who chose BPM, anxiety decreased significantly from 41/71 (58%) preoperatively to 29/71 (41%) 6 months post‐op ( p=0.04). and remained low at 18 months post‐op.

COMPARISON GROUP
Psychological morbidity showed a trend towards a decrease in the 64 women who declined BPM from baseline 57% (31/54) vs. 43% (23/54) at 6 months. % 14% (95% CI 0, 29) (p=0.08). Changes from baseline 57% (29/52) vs. 18 months 41% (21/52). % reduction=16% (95% CI‐2‐33) p=0.11.

CANCER WORRY
Significantly more women in the BPM group 24/74 or 32% compared to the No BPM group 6/58 or 10% were likely to believe that it was inevitable that they would develop breast cancer (P=0.03).

SEXUALITY
The degree of sexual pleasure did not change significantly in either group.

BODY IMAGE
Body image questionnaires given at the 6 and 18 month post‐operative interviews to accepters showed no difference in median score of 4 for body image on a scale of 0‐30 with 0 being the most positive

FOLLOWUP
Those choosing BPM were interviewed again at 6 and 18 months post‐operatively. Those declining or deferring were re‐interviewed at 18 months after the first interview

ATTRITION
11 /168 were lost to contact before completing assessment

Patients were assessed with 6 questionnaires measuring general health, anxiety, sexual activity, coping, risk perception and body image.

A score of 4 or higher on the General Health Questionnaire (GHQ) defined possible psychological morbidity.

Patients were identified from cohort of 168 women having a family history of breast cancer or having sufficiently high risk factors for BPM to be offered. They were followed prospectively with baseline data being collected prior to having BPM.

The comparison group is women who considered BPM, but declined. Of these, 154 were recruited for the study. Eleven deferred their decisions whose results were not reported.

Baseline statistical analysis included all women who completed the assessment at the first interview. In subsequent analyses, only those women who completed assessments at each time point are included.

Most women in both groups were employed and had children.

The median age of accepters was 38 and for decliners, 40.

Psychological morbidity decreased significantly over time among accepters, and the longer the time from surgery, the greater the decline.29% of the accepters had genetic testing vs. 5% of the decliners.

Frost 2000

Bilateral Prophylactic Mastectomy (BPM)

DECISION SATISFACTION
393/562 (70%) were either satisfied or very satisfied with their PM,
69/562 (11%) were neutral
107/562 (19%) were dissatisfied or very dissatisfied.
(Data missing from 10 participants.)

383/572 (67%) indicated they would definitely or probably choose PM again.

There was correlation between lower level of satisfaction and physician's advice being given as the primary reason for choosing PM.

CANCER WORRY
423/572 (74%) reported a diminished level of emotional concern about developing breast cancer.
520/572 (91%) of the women reported no change or favorable effect on emotional stability.
52/572 (9%) reported adverse effect in level of emotional stability.
492/572 (86%) of the women reported no change or favorable effect on stress levels.
80/572 (14%) reported adverse effect in level of stress.

BODY IMAGE
275/572 (48%) reported no change in their level of satisfaction with their physical appearance.
92/572 ( 16%) reported favorable effects.
206/572 (36%) reported diminished satisfaction with their physical appearance.

429/572 (75%) of the women reported no change or favorable effect in feelings of feminity.
132/572 (23%) reported adverse effect in feelings of femininity.

Variable most strongly associated with patient satisfaction after BPM was satisfaction with body appearance

469/572 (82%) of the women reported no change or favorable effect in self‐esteem
103/572 (18%) reported adverse effect in self‐esteem.

SEXUALITY
440/572 (77%) of the women reported no change or favorable effect.
132/572 (23%) reported adverse effect in sexual relationships.

FOLLOWUP
Mean follow‐up period was 14.5 years after surgery.

ATTRITION
572 of 609 (94%) completed the questionnaire

Patient satisfaction assessed by questionnaire to evaluate long‐term satisfaction, and psychological and social function.

The 609 women were a subset of 639 subjects in Hartmann's 1999 study known to be alive and were recruited to complete a study questionnaire after their BPM to evaluate their long‐term satisfaction, and psychological and social function.

572 of 609 (94%) completed the questionnaire.

Family history was the most common number 1 reason given for having a BPM, followed by physicians' advice and nodular breasts. Because reason for choosing BPM was not collected pre‐operatively, authors are concerned that recall of reason for choosing BPM may have been colored by subsequent experience.

100% of the 19 women who did not have reconstruction reported being very satisfied or satisfied, and using multiple regression analysis showed there was an association between satisfaction and no reconstruction

Josephson 2000

Bilateral Prophylactic Mastectomy (BPM)

DECISION SATISFACTION
No women regretted their decision to have PM. They believed they had no choice in the decision since minimizing risk of breast cancer was most important.

All women would recommend BPM with immediate reconstruction to other women in the same situation.

COSMETIC SATISFACTION
3/15 (87%)women thought the cosmetic result was better than expected.

8/15 (53%) did not feel their new breasts were part of their own body.

MEDICAL CARE SATISFACTION
14/15 (93%) women were satisfied with the factual information given by the geneticists.
8/15 (53%) women were satisfied with the surgical/reconstruction information.
9/15 (60%) were satisfied with the support from their surgeons.

SEXUALITY
5/13 (38%) women with partners said that the results of the operation had changed the relationship with the partner.

FOLLOWUP
13 interviews took place at least 1 year after surgery, 2 at less than 1 year (7 and 8 months).

ATTRITION
2/17 women selecting BPM refrained from being interviewed.

Quality of Life satisfaction with BPM as measured by post‐operative interviews with 17 women.

Satisfaction with PM as measured by post‐operative interviews.

5 of the women spontaneously mentioned the importance of how to emotionally consider and anticipate the loss of their breasts, how the breasts would be changed following the surgical procedure and "taking good‐bye" of them, mourning them.

Lloyd 2000

Bilateral Prophylactic Mastectomy (BPM)

Results produced 7 significant categories, 6 stages and a seventh‐suffering and countering multiple losses‐which transcended all 6 stages.

The six stages are:
1.deciding
2.telling
3.experiencing surgery and recovery
4.maintaining womanliness
5.processing the loss
6.moving on

FOLLOWUP
Participants were between 6 weeks and 3 years post surgery.

ATTRITION
1/ 11 subjects refused to participate

Qualitative in‐depth interviews were conducted to explore personal experiences with prophylactic mastectomy and ascertain categories or stages women go through who have BPM.

Hopwood 2000

Bilateral Prophylactic Mastectomy (BPM)

CANCER WORRY
47 of 49 returned evaluable General Health Questionnaires (GHQ) one‐year post‐operatively.

8/47 (17%) scored >9 in a range of 0‐28 suggesting "case" level distress.

BODY IMAGE
All 49 returned Body Image Scale (BIS) questionnaires one year post‐operatively
.6/49 (12%) reported moderately changed or very much change overall in body image on 10 items.

More than half of the women reported a change from little to very much for 3 items.
27/49 (55.1%) felt less sexually attractive
26/49 (53.1%) felt self‐conscious about appearances
26/49 (53.1%) feel less physically attractive.

FOLLOWUP AND ATTRITION

19/49 women had 1‐ and 2‐ year assessments.9 /49 had 1‐, 2‐ and 3‐year assessme

Quality of Life measured by General Health Questionnaire (GHQ) and Body Image Scale (BIS) to assess mental health and body image one year post‐operatively.

Participants were recruited from a group of 76 women who had BPM7 of 45 women required further psychiatric help.

3 of the 7 were given antidepressant medication.

Complications from surgery accounted for 4 of the 7 women needing psychiatric help.

Surgical complications e.g. skin necrosis, nipple loss, infection and pain, accounted for some of the highest GHQ and BIS scores.

Montgomery 1999

Contralateral Prophylactic Mastectomy (CPM)

DECISION SATISFACTION
Most women were satisfied with CPM.18 of 296 (6%) regretted their decision to have CPM with 11/296 (5%) of them among the 212 who said the discussion about CPM was initiated by the physician.

COSMETIC SATISFACTION
12/111 who had reconstruction had regrets
6/185 who did not have reconstruction had regrets. RR=0.30 (95% CI 0.12, 0.78) (p=0.01)

88/111 (79%) who underwent reconstruction reported their cosmetic results were excellent or acceptable.
18/111 (16%) said cosmetic results were unacceptable, but only 12 of them also had regrets.
5/111 (5%) did not report satisfaction.

6/111 (5.4%) said they would not chose CPM again if they had known the cosmetic outcome.

SEXUALITY
The reasons given by the 18 women with regrets were:
7/18 (39%) cosmetic results
4/18 (22%) diminished sense of sexuality
4/18 (22%) lack of education about alternatives
3/18 (17%) other reasons.

FOLLOWUP
The follow‐up years since surgery ranged from 0.25‐43.8 years with a mean of 4.9 years.

ATTRITION
50 women of 346 did not respond to questionnaire

Quality of Life/Satisfaction assessed by survey regarding satisfaction and regrets with PM.

346 patients were selected from a group of 817 volunteers who responded to an invitation in the popular press to join the National Prophylactic Mastectomy Registry and who had CPM.

Insurance companies overwhelmingly provided coverage for CPM in 276 women (93%).

Regrets were less common, but not statistically significant, among women with whom the discussion to have CPM was initiated by the physician (11/212 or 5%) than among women who initiated the discussion themselves 7/84 or 8%).

Study

Survival

Followup

Attrition

Study Details

Babiera 1997

5‐year disease‐specific survival
CPM group= 80% (14/18)
N o CPM (Surveillance) = 90% (104/115)
RR=0.87 (95% CI 0.67, 1.12) p=0.28

CPM group
Median =52 months
(range 17‐143 months).

No CPM group
Median = 70 months (range 13‐178 months).

None

The authors conclude that the results suggest that CPM should not be considered as an initial therapeutic option in patients presenting with ILC.

In the CPM group, 39% of the patients had adjuvant chemotherapy, 6% underwent only neoadjuvant therapy, 28% had radiation therapy, and 22% had hormonal therapy as treatments for their primary ILC.

In the COMPARISON group, 31% had adjuvant therapy, 8% had both adjuvant and neoadjuvant therapy, and 28% underwent hormonal therapy as treatment for their primary ILC.

Peralta 2000

15‐year all‐cause survival
CPM group = 64% (41/64)
No CPM group = 48% (87/182) (P=0.26)

15‐year disease‐specific survival in pts with Stage 0, 1 or 2 breast cancer
CPM group =71% (45/64)
No CPM (Surveillance) = 53% (96/182) (P=0.06).

CPM group:
Median = 6.2 years

No CPM group:
Median= 6.8 years

None

Comparison group patients were matched for age, stage of disease at diagnosis, presence of LCIS, chemotherapy and tamoxifen therapy from among 2,852 patients who underwent mastectomy between 1/1/1973 and 9/30/1998 at one institution.

71% having CPM had immediate reconstruction.

Lee 1995

15‐year disease‐specific survival
CPM or biopsy =105 patients
No CPM (Surveillance)=299

There was a statistically significant 15‐year survival advantage in CPM or biopsy (P=0.01) after adjusting for age.

Mean = 6 years
Median= 5.3 years

None

Patients had unilateral invasive lobular carcinoma (ILC)

Patients in the CPM group were significantly younger and a significantly greater proportion had multifocal lesions than in the No CPM group. Results were age adjusted.

Those getting CPM and those only getting biopsies were lumped together the 'treatment group.' There are no statistical analyses of just the CPM group alone.

Study

Incidence

Follow‐up Time

Attrition

Study Details

Leis 1981

At 10 years:
CPM …………….4/58

No comparison group

Mean = 6 years Median= 5.3 years

69/127 (54%) subjects not accounted for at 10 years

25 of 127 (19.7%) had unsuspected cancer in the contralateral breast at the time of CPM; 11 were invasive and 14 were non‐invasive.

Babiera 1997

TREATMENT GROUP:
CPM=not reported

COMPARISON GROUP:
No CPM (Surveillance)=3/115 (2.6%)

CPM
Median= 52 months
(range 17‐143 months).

No CPM
Median = 70 months (range 13‐178 months).

None

The authors conclude that the results suggest that CPM should not be considered as an initial therapeutic option in patients presenting with ILC.

In the CPM group, 39% of the patients had adjuvant chemotherapy, 6% underwent only neoadjuvant therapy, 28% had radiation therapy, and 22% had hormonal therapy as treatments for their primary ILC.

In the COMPARISON group, 31% had adjuvant therapy, 8% had both adjuvant and neoadjuvant therapy, and 28% underwent hormonal therapy as treatment for their primary ILC.

Peralta 2000

TREATMENT GROUP: CPM…………0/64

COMPARISON GROUP:
No CPM (Surveillance) …..36/182
RR=0.04 (95% CI 0.00, 0.62) p=0.02

CPM group
Median = 6.2 yrs
No CPM group
Median= 6.8 yrs

None

Comparison group patients were matched for age, stage of disease at diagnosis, presence of LCIS, chemotherapy and tamoxifen therapy from among 2,852 patients who underwent mastectomy between 1/1/1973 and 9/30/1998 at one institution.

71% had immediate reconstruction.

McDonnell 2001

Premenopausal Women:
TREATMENT GROUP:
CPM=6/388

COMPARISON GROUP:
Statistically simulated, adjusted for treatment= 106.2/388
Risk reduction =94.4% (95% C.I. 87.7% ‐ 97.9%).

Not adjusted for treatment=115/388
Risk reduction=94.8% (95% C.I. 88.6% ‐ 98.1%)

Postmenopausal Women
TREATMENT GROUP:
CPM=2/357

COMPARISON GROUP
Statistically simulated, adjusted for treatment=50.3/357
Risk reduction = 96.0% (95% C.I. 85.6% ‐ 99.5%).

Not adjusted for treatment =54/357 96.3%.

Median = 10 years

98% of participants were followed at least 2 yrs

None

Total occurances of contralateral cancers among all women was 8/745

742 of the subjects fit the Anderson model definition of positive family history, which requires one of the three types of pedigrees: parent affected, sibling affected or second‐degree relative affected.

3 women who developed contralateral breast cancer after CPM and whose family pedigree was unclear were included to make the calculated risk reductions conservative.

The median time from mastectomy to development of breast cancer was 2 years (range 1‐18 years).

"Adjusted for treatment" means adjusted for adjuvant therapy and Tamoxifen

Comparison group was statistically simulated using age‐adjusted life tables.

4 of the cancers were diagnosed within 2 years of CPM, suggesting that the cancer may have been present but not detected at that time.

Study

Outcome

Followup Time

Attrition

Study Details

Peralta 2000

15‐year disease‐Free Survival (Recurrence or secondary primary in contralateral or primary):

TREATMENT GROUP:
CPM group = 55% (35/64)

COMPARISONG GROUP
No CPM = 28% (51/182) (P=0.01).

CPM group:
Median = 6.2 years

No CPM group:
Median= 6.8 years

None

Comparison group patients were matched for age, stage of disease at diagnosis, presence of LCIS, chemotherapy and Tamoxifen therapy from among 2,852 patients who underwent mastectomy between 1/1/1973 and 9/30/1998 at one institution.

71% had immediate reconstruction.

Leis 1981

Disease‐free survival at 10 or more years

TREATMENT GROUP
CPM...............................54/58 (91%)

No comparison group

Mean = 6 years Median= 5.3 years

69/127 (54%) subjects not accounted for at 10 years

25 of 127 (19.7%) had unsuspected cancer in the contralateral breast at the time of CPM; 11 were invasive and 14 were non‐invasive.

Babiera 1997

RECURRENCE

Local ‐regional recurrences
TREATMENT GROUP
CPM……………….. 0/18

COMPARISON GROUP
No CPM (Surveillance): ..10/115
RR=0.29 (95% CI 0.02, 4.76), p=0.39

Distant disease
TREATMENT GROUP
CPM………………..3/18

COMPARISON GROUP
No CPM (Surveillance)………25/115
RR=0.77 (95% CI 0.26, 2.28) p=0.63

CPM group
Median =52 months
(range 17‐143 months).

No CPM group
Median = 70 months (range 13‐178 months).

None

In the CPM group, 39% of the patients had adjuvant chemotherapy, 6% underwent only neoadjuvant therapy, 28% had radiation therapy, and 22% had hormonal therapy as treatments for their primary ILC.

In the COMPARISON group, 31% had adjuvant therapy, 8% had both adjuvant and neoadjuvant therapy, and 28% underwent hormonal therapy as treatment for their primary ILC.