Int. J.

Cancer: 98, 596 – 603 (2002) Publication of the International Union Against Cancer

© 2002 Wiley-Liss, Inc.

DOI 10.1002/ijc.10194
BRCA1 AND BRCA2 MUTATIONS AMONG BREAST CANCER PATIENTS
FROM THE PHILIPPINES
Maria Lourdes DE LEON MATSUDA1, Alexander LIEDE2*, Elaine KWAN2, Cynthia A. MAPUA1, Eva Maria C. CUTIONGCO3, Alex TAN4,
Åke BORG5 and Steven A. NAROD2
1
Department of Surgery, College of Medicine and Philippine General Hospital, University of the Philippines, Manila, Philippines
2
Centre for Research in Women’s Health, University of Toronto, Canada
3
Institute of Human Genetics, National Institutes of Health, University of the Philippines, Manila, Philippines
4
Department of Surgery, Davao Regional Hospital, Tagum City, Philippines
5
Department of Oncology, University Hospital, Lund, Sweden

Age-adjusted incidence rates of breast cancer vary more single BRCA2 mutation (BRCA2 999del5) is found in 24% of
than 10-fold worldwide, with the highest rates reported in women diagnosed with breast cancer below age 40.9,19 Among
North America and Europe. The highest breast cancer inci- unselected Jewish breast cancer cases, 12% were found to carry 1
dence rates in Southeast Asia have been reported for the of 3 founder BRCA mutations, including 28% of those diagnosed
Manila Cancer Registry in the Philippines, with an age-stan- below age 50.20 In general, there is little data on the contribution
dardized rate of 47.7 per 100,000 per year. The possible
contribution of hereditary factors to these elevated rates has of germline BRCA1 and BRCA2 mutations to breast cancer in East
not been investigated. We conducted a case-control study of Asia. For example, there has only been a case-report of a BRCA1
294 unselected incident breast cancer cases and 346 female mutation (C2178T) in a Filipino breast-ovarian patient, residing in
controls from Manila, Philippines. Cases and controls were the continental United States.21 One of the characteristics of he-
selected from women below the age of 65 undergoing eval- reditary breast cancer is a tendency towards younger age at onset.
uation at the PGH in Manila because of a suspicious breast Breast cancer is typically diagnosed at a young age in Southeast
mass. Molecular analysis identified 12 BRCA2 mutations and 3 Asia, suggesting that genetic factors may be important. We present
BRCA1 mutations. We estimate the prevalence of BRCA mu- family history data and molecular analyses from a case-control
tations among unselected breast cancer cases in the Philip-
pines to be 5.1% (95% CI: 2.6 –7.6%), with a prevalence of 4.1% study of unselected incident breast cancer patients from the Phil-
(95% CI: 1.8 – 6.4%) for BRCA2 mutations alone. The BRCA2 ippines.
4265delCT and 4859delA mutations were found in 2 and 4
unrelated cases, respectively; haplotype analysis confirmed
that these, and the BRCA1 5454delC mutation, are founder METHODS
mutations. BRCA2 mutations were also found in 2 of 346 Subjects and study design
controls (0.6%; 95% CI: 0.2–1.4%). Compared with non-car-
rier cases, the cumulative risk of breast cancer for first- The Philippine General Hospital (PGH) serves the metropolitan
degree relatives of mutation carriers was 24.3% to age 50, Manila area. This densely populated region consists of 5 cities
compared with <4% for first-degree relatives of non-carrier (Manila, Makati, Quezon City, Pasig and Caloocan ) and 13 towns
cases (RR ⴝ 6.6; 95% CI: 2.6 –17.2; pⴝ 7.5 ⴛ 10ⴚ6). Our data or municipalities, with a total population exceeding 10 million
suggest that penetrance of BRCA mutations is not reduced in inhabitants. The region has an annual average of over 9,000 new
the Philippines. Germline mutations in the BRCA2 gene con- cancer cases a year.22 The PGH provides oncology services to
tribute more than mutations BRCA1 to breast cancer in the 20 –25% of the population of metropolitan Manila and Rizal prov-
Philippines, due in large part to the presence of 2 common
founder mutations. ince. The PGH is a charity hospital and receives referrals from all
© 2002 Wiley-Liss, Inc. regions of the Philippines as the primary government-sponsored
tertiary care facility.
Key words: BRCA1, BRCA2, Philippines; breast cancer; Asia; un- Breast cancer cases were ascertained in the course of a case-
selected control study conducted at the PGH in Manila, from August 1997
to June 2000; active case recruitment occurred during a 1-year
period in 1998. This collaborative research project was designed to
The incidence of breast cancer varies as much as 10-fold be- explore possible genetic and non-genetic risk factors responsible
tween countries.1 Breast cancer incidence rates in North America, for the high risk of breast cancer in the Philippines. Questionnaire
Western Europe and Scandinavia are much higher than rates in data and blood specimens were collected on 294 incident breast
Southeast Asia. Studies of Japanese and Chinese migrants to North cancer cases. Subjects were interviewed at the outpatient
America show that the incidence rates of breast cancer in migrants breast clinic of the PGH in Manila at the time of evaluation of a
assume the rate in the host country within 1–2 generations, impli- breast complaint, i.e., the breast complaint was usually a breast
cating environmental factors. The Philippine Manila Cancer Reg-
istry has reported the highest incidence rates of breast cancer in
Asia.1 The (world-standardized) rate of 47.7 per 100,000 exceeds Grant sponsor: Canadian Breast Cancer Foundation.
the rate reported for several Western countries, including Spain,
Italy and most Eastern European countries.
Following the identification of the breast and ovarian cancer The first two authors contributed equally to ths study.
susceptibility genes BRCA1 (MIM 113705) and BRCA2 (MIM
600185), the frequency and spectrum of disease-related mutations *Correspondence to: Centre for Research in Women’s Health, Univer-
has been investigated in various geographic regions and ethnic sity of Toronto, 790 Bay Street, Suite 750A, Toronto, Ontario M5G 1N8,
groups (Breast Cancer Information Core at [Link] Canada. Fax: ⫹416-351-3767. E-mail: [Link]@[Link]
gov/Intramural_research/Lab_transfer/Bic/). The contribution of
BRCA1 and BRCA2 mutations to breast cancer rates has been
Received 14 February 2001; Revised 14 August, 11 October 2001;
examined in North America,2– 6 United Kingdom,7,8 Iceland,9,10 Accepted 19 October 2001
Spain,11,12 the Netherlands,13 Hungary,14 Germany,15 Finland16
and Australia.17,18 In several ethnic groups, a number of BRCA1
and BRCA2 founder mutations have been identified. In Iceland, a Published online 30 January 2002
 BRCA1 AND BRCA2 MUTATONS IN THE PHILIPPINES 597
mass but also included women with breast pain or nipple dis- data and Student’s t-tests were used for comparison of continuous
charge. Approximately 800 incident cases of breast cancer are variables.
diagnosed annually at the PGH. Cases included a sample of Data from the Manila cancer registry as published in Cancer
women aged 25– 65 years with a histologically confirmed diagno- Incidence in Five Continents vol. VII1 was used as an alternate
sis of a first primary invasive breast cancer. These cases represent source of controls for analysis of cancer risk. The expected number
those presenting at a single weekly outpatient clinic; although this of cases was calculated from the product of the person-years and
is a relatively small proportion of total number of cases, we believe the Manila cancer registry age-specific cancer rates. The relative
that they are representative of the larger hospital patient popula- risks (RR) of cancer in first-degree relatives was estimated by
tion. Our control group comprised of 346 women without breast comparing observed with expected figures. The confidence inter-
cancer and included 229 women referred to the same breast clinic vals (CI) were calculated assuming a Poisson distribution.26
for evaluation, but who did not receive a biopsy (n⫽66), or who
had a negative biopsy (n⫽163). Additional non-cancer controls
included 117 female patients attending outpatient clinics at the RESULTS
PGH for other conditions. Compliance was 97% for cases and 95% The characteristics of the 294 incident breast cancer cases from
for controls. the breast clinic of the Philippine General Hospital are summarized
in Table I. The mean age of diagnosis was 44.0 years (range 25– 63
Mutation analysis years). The mean age of the controls was 42.2 years (range 25– 64
Human genomic DNA was isolated from 20 mL of peripheral years). Fifteen germline BRCA1/2 mutations were detected among
blood. Blood specimens were sent to the molecular laboratory of the 294 incident breast cancer cases from Manila (5.1%; 95% CI:
the Sunnybrook and Women’s College Health Sciences Centre in 2.6 –7.6%) (Table II). The previously reported BRCA1 exon 11
Toronto, Canada. A variety of methods were employed to detect
the presence of a BRCA1 or BRCA2 mutation. Exon 11 of BRCA1
and exons 10 and 11 of BRCA2 were screened by the protein- TABLE I – CHARACTERISTICS OF CASES AND CONTROLS
truncation testing (PTT) for all cases. All mutant bands detected by FROM MANILA, PHILIPPINES
PTT were confirmed with direct sequencing. Seven cases with 2 or BRCA1/2
Characteristics
more relatives with breast (below age 50) or ovarian cancer were Cases (%) Cases (%) Controls (%)
selected for screening by direct sequencing of all coding regions
through Myriad Genetic Laboratories ([Link] Total (n) 294 15 346
Age at diagnosis (years)
Any mutation detected by Myriad Genetic Laboratories was ana- 25–34 34 (12) 2 (13) 80 (23)
lyzed for all study subjects. Furthermore, we tested all study 35–44 127 (43) 8 (53) 138 (40)
subjects for the BRCA1 exon 22 5454delC mutation, detected in a 45–54 114 (39) 5 (33) 106 (31)
Filipino breast cancer patient from Hawaii (personal communica- 55–64 19 (6) — 22 (6)
tion). Analysis of BRCA1 exons 15, 22 and 24 was performed by Family history
denaturing gradient gel electrophoresis (DGGE)23 PCR primers (first-degree relatives)
were fluorescently-labeled and the PCR fragments were visualized Breast cancer 31 (11) 5 (33) 32 (9)
with the use of a phosphorimager. In summary, we screened Breast cancer (⬍50 years) 19 (7) 5 (33) 17 (5)
BRCA1 exons 11, 15, 22, 24 and BRCA2 exons 10 and 11 for all Breast cancer (2 or more) 6 (2) 2 (13) 4 (1)
Male breast cancer 2 (1) 2 (13) 0
breast cancer cases. This testing is estimated to detect 60 – 80% of Ovarian cancer 5 (2) 2 (13) 1 (⬍1)
all germline mutations in the coding region of these 2 genes.8 Unknown 16 (5) — 73 (21)
Control subjects were tested for the recurrent mutations in exon Tumor histology
11 of BRCA2 and exons 15 and 22 of BRCA1. Infiltrating ductal 257 (87) 15 (100)
Infiltrating lobular 3 (1) —
Genotyping markers in the BRCA1 and BRCA2 regions Papillary 7 (2) —
Medullary 3 (1) —
Three mutations were found in multiple patients: BRCA2 Other 18 (6) —
4265delCT, 4859delA and the BRCA1 5454delC. To investigate Unknown 6 (2) —
whether the BRCA2 4265delCT and the 4859delA mutations are Stage
associated with specific genotypes, we analyzed 6 polymorphic I 8 (3) —
microsatellite markers located at the BRCA2 locus on chromo- II 95 (32) 4 (27)
somes 13q (D13S1699, D13S1698, D13S1697, D13S1701, III 154 (52) 11 (73)
D13S171 and D13S1695). The BRCA1 5454delC mutation was IV 31 (10) —
examined using the 3 BRCA1 intragenic microsatellite markers Unknown 6 (2) —
Size
D17S1323, D17S1322 and D17S855. Fluorescence-labeled PCR T1 11 (4) —
products were run in an ABI310 DNA analyzer and evaluated T2 93 (32) 7 (47)
using Genescan software (Applied Biosystems, Foster City, CA) T3 64 (22) 3 (20)
and a fluorescence-labeled DNA fragment size standard. Primer T4 121 (41) 5 (33)
sequences to amplify these markers can be retrieved on-line from Unknown 5 (2) —
the GDB (Human Genome Database). A physical map for 13q and Estrogen receptor1
17q markers has been previously resolved.24,25 Negative 52 (41) 3 (38)
Positive 76 (60) 5 (62)
Statistical analysis Nuclear grade1
I 15 (9) 1 (12)
The observed number of cases in the first-degree relatives was II 80 (49) 4 (50)
determined by review of family pedigrees. Kaplan-Meier survival III 67 (41) 3 (38)
analysis was used for the calculation of the cumulative incidence Lymph node involvement1
of cancer in first-degree relatives of cases and controls. First- Yes 191 (67) 7 (50)
degree female relatives were considered to be at risk of cancer No 93 (33) 7 (50)
from birth until death, or age at date of interview with proband. Distant metastases1
The log-rank test used to assess statistical significance of differ- Yes 33 (12) 1 (7)
No 255 (88) 14 (93)
ences in survival curves. The relative risks (RR) of cancer among
first-degree relatives were estimated by use of the Cox propor- 1
Percentages were calculated from cases for whom this information
tional hazards model. Fisher’s exact tests were used for nominal was available.
 TABLE II – CHARACTERISTICS OF BRCA1 AND BRCA2 MUTATION CARRIERS FROM THE PHILIPPINES

Case/ Age Family history of

Subject Exon mutation Family history of other cancers Region in Philippines
Control (years) breast/ovarian cancer

In study BRCA1 GENE

1 PH0199 15 Q1538X Case 45 Br 70s, Ov 50s, Br 40s Re 80, Li 60s, Li 50s Bulacan (Central Luzon)
2 PH0047 22 5454delC Case 42 Nil Nil Marinduque (Southern Tagalog)
3 PH0019 24 R1835X Case 36 Br 45, Br 45, Br 40s Nil Nueva Ecija (C. Luzon)
BRCA2 GENE
1PH0052 11 4265delCT Case 49 Nil Li 40s Cavite (S. Tagalog)
2PH0150 11 4265delCT Case 44 Nil Nil Samar (Eastern Visayas)
3PH0044 10 2042insA Case 36 Nil Ga 40s Laguna (Southern Tagalog)
4PH0186 11 3827delGT Case 32 Br 39 Nil Masbate (Bicol)
5PH03621 11 C3590G Case 43 HBOC Br 27, Bt? Aklan (Western Visayas)
6PH03651 11 C3590G Case 45 HBOC Bt 27, Bt? Aklan (W. Visayas)
7PH0392 11 3798delG Case 47 Br 30s Nil Albay (Bicol)/Davao/Nueva Ecija (C. Luzon)
8PH0744 11 6083insAGTT Case 26 Br 27, Br 70 Nil Iloilo (W. Visayas)/Mindoro (S. Tagalog)
9PH0153 11 4859delA Case 48 Nil Nil Leyte (E. Visayas)
10PH0477 11 4859delA Case 41 Br 52 Lu 65, Lu 62, Pa 73, Lu 51 Boracay (W. Visayas)
11PH0508 11 4859delA Case 50 Nil Nil Leyte (E. Visayas)
12PH0563 11 4859delA Case 36 Br 43, Br 58 Ut 73 Pampanga (C. Luzon)
PH0307 11 3827delGT Control 56 Nil Li ?, Or? Caloocan (Greater Manila)
PH0326 11 5168delCA Control 32 Br 44 Pr 79, Re 54 Laguna (S. Tagalog)
Outside Study BRCA1 GENE
PH05612 22 5454delC — 30 HBOC Pr 67 Cavite (S. Tagalog)
BRCA2 GENE
3
PH0356 11 4859delA — 15 Br 33 Pr 60s (Greater Manila)
PH03784 11 4859delA — 44 Nil Nil Leyte (E. Visayas)
FOTS102855 11 4859delA — 49 Br 45 Li ?, Lu ? N/A
PH00936 11 4265delCT — 31 Br 53 Li 60, Li 60 Bulacan (S. Tagalog)
Patients indicated include–12 sisters with incident breast cancers,–2clinic referral,–3patient with childhood-onset non-Hodgkin’s lymphoma,–4ovarian cancer patient with benign breast disease
(biopsy negative),–5patient ascertained from study of incident ovarian cancer in Ontario27 and a–6prevalent breast cancer patient. Abbreviations for family members’ type of cancer: Br ⫽ breast
cancer, Ov ⫽ ovarian cancer, Li ⫽ liver cancer, Bt ⫽ malignant brain tumor, Re ⫽ rectal cancer, Ga ⫽ gastric cancer, Pa ⫽ pancreatic cancer, Lu ⫽ lung cancer, Ut ⫽ uterine cancer, Or ⫽
oral cancer, HBOC ⫽ hereditary breast-ovarian cancer syndrome, 3⫹ breast (⬍50 years) or ovarian cancer.
 BRCA1 AND BRCA2 MUTATONS IN THE PHILIPPINES 599

mutation (C2178T) in a Filipino breast-ovarian patient residing in (PH0362 and PH0365). Seven BRCA2 mutations were identified, 2
the United States was not detected among our series of patients in multiple patients. The BRCA2 4265delCT mutation was found
from the Philippines.21 There were 12 mutations in the BRCA2 in 2 breast cancer cases and the BRCA2 4859delA was found in 4
gene (4.1%; 95% CI: 1.8 – 6.4%), accounting for 80% of all mu- cases. Seven of the 11 unrelated carriers of mutations had a family
tations detected. The rate of mutation detection was higher among history of breast or ovarian cancer (Table II).
breast cancer cases diagnosed below age 45 (6.2%; 95% CI: Two control subjects were found to carry exon 11 BRCA2
2.4 –10%) than among those diagnosed at age 45 or older (3.8%; mutations (0.6%; 95% CI: 0.2–1.4%); both women were clinic
95% CI: 0.5–7%) (Table III); no mutations were observed in controls (Table II). A 56-year-old control subject (PH0307) was
women diagnosed at age 55 and older. The mean age of breast found to carry the BRCA2 exon 11 3827delGT mutation — a
cancer diagnosis among BRCA1/2 mutation carriers was 42.1 years mutation found in also detected among our cases (PH0186). This
and was not statistically different from the mean age of 44.2 years woman attended the Philippine General Hospital breast clinic for
for the mutation-negative cases (p⫽ .28). The clinical character- diagnostic mammography, interpreted as fibrocystic disease, and a
istics of BRCA2-related breast cancers are summarized in Table IV biopsy was not recommended. A second 32-year-old control sub-
and compared with mutation-negative cases. Twelve BRCA2 mu- ject (PH0326) attended the clinic on several occasions and under-
tations were detected in 11 different families; 2 sisters with the went multiple excision biopsies, which exhibited fibrocystic dis-
BRCA2 exon 11 C3590G mutation were ascertained independently ease. Her most recent biopsy revealed a benign lesion of the breast
(squamous papilloma). Pedigrees of the mutation-positive controls
TABLE III – FREQUENCY OF BRCA1 AND BRCA2 MUTATIONS are presented in Figure 1. Eighty percent of the Manila breast
AMONG BREAST CANCER CASES FROM THE PHILIPPINE cancer cases provided sufficient detail (i.e., age and type of cancer
GENERAL HOSPITAL, MANILA
diagnosis) on first-degree relatives for familial cancer risk estima-
Age of Number of BRCA1 BRCA2 Either tion. Information on family history was also reviewed with the
diagnosis mutation mutation mutation
(years) cases number (%) number (%) number (%) mothers of the cases and controls where possible. Seventy-nine
percent (140/178) of living mothers of cases and 87% (198/227) of
25–34 34 0 (0) 2 (5.9) 2 (5.9) living mothers of controls were interviewed. The relatives of the 2
35–44 127 3 (2.4) 5 (3.9) 8 (6.3) sisters who were ascertained separately (PH0362 and PH0365)
45–54 114 0 (0) 5 (4.4) 5 (4.4) were included in survival analysis only once. This was also the
55–64 19 0 (0) 0 (0) 0 (0)
All ages 294 3 (1) 12 (4.1) 15 (5.1) only family with a case of male breast cancer (Fig. 1).
The cumulative incidence of breast cancer among female first-
degree relatives of cases was not statistically greater than that
TABLE IV – CANCER INCIDENCE AMONG FIRST-DEGREE among control relatives. Using published data from the Manila
RELATIVES OF CASES
Cancer Registry, the risk of cancer was significantly increased for
Relatives of first-degree relatives of cases compared with the general popula-
Age 294 cases Expected1 RR (95% CI)
observed tion. This data is presented in Table IV.
Seven of the 15 cases with a germline mutation had a first or
Any cancer Any 98 83.8 1.2 (0.95–1.4)
⬍50 50 65.5 0.76 (0.57–1) second-degree relative with breast cancer or ovarian cancer com-
ⱖ50 48 30.8 1.6 (1.2–2.1) pared with 51 of 264 non-carriers (odds ratio 3.65; 95% CI:
Any cancer Any 66 48.4 1.4 (1.1–2.7) 1.3–10.5; p ⫽ .019). The cumulative incidence of cancer among
(female) ⬍50 35 20.2 1.7 (1.2–2.5) the first-degree relatives of carriers of mutations was significantly
ⱖ50 31 31.3 1.0 (0.67–1.4) greater than that among the relatives of the mutation-negative
Breast cancer Any 37 12.2 3.0 (2.1–4.2) cases (P ⫽ 7.2 ⫻ 10⫺5). The lifetime RR for breast cancer in
(female) ⬍50 23 5.7 4.0 (2.6–6.1) female first-degree relatives was 5.0 (95% CI: 2.0 –12.0) compared
ⱖ50 14 7.1 2.0 (1.1–3.3) with relatives of mutation-negative cases. This effect was stronger
1
From Cancer Incidence in Five Continents vol. VII1 for the risk of breast cancer before age 50 (RR ⫽ 6.6; 95% CI:

FIGURE 1 – Pedigrees of selected cases and

controls with BRCA1/2 mutations. Circles indi-
cate women and squares indicate men. Black
circles indicate women affected with either breast
or ovarian cancer. Black squares indicate men
affected with breast cancer. Gray circles and
squares indicate individuals affected with cancers
other than breast or ovarian cancer. Diagonal
slash indicates deceased. Ov ⫽ ovarian cancer,
Br ⫽ breast cancer, Pr ⫽ prostate cancer, Pa ⫽
pancreatic cancer, Bt ⫽ malignant brain tumor,
Li ⫽ liver cancer, Lu ⫽ lung cancer, Or ⫽ oral
cancer, Ut ⫽ uterine cancer, Re ⫽ rectal cancer,
Psu ⫽ primary site of cancer unknown. The num-
bers following the abbreviations indicate age of
diagnosis. The plus sign indicates the presence of
a germline BRCA1 or BRCA2 mutation (Table
II).
600 DE LEON MATSUDA ET AL.

FIGURE 1 – CONTINUED.

TABLE V – GENOTYPES FOR BRCA1 OR BRCA2 MUTATIONS IDENTIFIED IN MULTIPLE SUBJECTS1

BRCA2
Family Ancestry Region D13S1699 D13S1698 D13S1697 D13S1701 D13S171 D13S1695
EXON11

Lund 53 Swedish Skane 151兩157 167兩167 225兩229 4265delCT 293兩305 230兩234 254兩256
PH0093 Hispanic/Chinese Bulacan (S. Tagalog) 157兩160 167兩171 221兩221 4265delCT 297兩313 232兩234 254兩254
PH0052 Malay/Filipino Cavite (S. Tagalog) 157兩160 165兩173 221兩225 4265delCT 301兩301 232兩234 252兩254
PH0150 Malay Samar (E. Visayas) 160兩160 167兩167 221兩225 4265delCT 301兩301 232兩234 252兩252
PH0153 Malay Leyte (E. Visayas) 173兩177 221兩225 4859delA 301兩301 230兩234
PH0378 Malay Leyte (E. Visayas) 167兩177 221兩221 4859delA 293兩301 230兩234
PH0477 Malay Borocay (W. Visayas) 169兩177 221兩225 4859delA 305兩309 230兩230
PH0508 Malay Leyte (E. Visayas) 171兩177 221兩225 4859delA 293兩305 230兩230
PH0563 Malay Papanga (Central Luzon) 173兩177 221兩225 4859delA 305兩309 234兩244
PH0356 Malay/Chinese Greater Manila 163兩177 221兩221 4859delA 289兩305 230兩230
FOTS10285 Filipino Greater Manila 167兩175 221兩221 4859delA 305兩313 230兩234
BRCA1 INTRON 19 INTRON 20
Family Ancestry Region INTRON 12 EXON 22
D17S1322 D17S855
D17S1323

PH0561 Malay/Hispanic Cavite (S. Tagalog) 158兩160 119兩122 140兩150 5454delC

PH0047 Malay/Hispanic Marinduque (S. Tagalog) 158兩160 119兩119 150兩152 5454delC
1
Common haplotype shown in bold.

2.6 –17.2; p⫽ 7.5 ⫻ 10⫺6). The cumulative incidence of breast the mutation is possible. The Swedish 4265delCT carrier displayed
cancer in female first-degree relatives of BRCA1/2 mutation car- a different genotype from the Filipino cases.
riers was 24.3% to age 50, compared with ⬍4% among first-
degree relatives of non-carrier cases (Fig. 2). The BRCA2 4859delA mutation was identified in 4 unrelated
incident breast cancer cases. In addition, we identified 3 cases with
Only 3 of the mutations described in this study have been
reported in other populations (BRCA2 2042insA, 4265delCT and the BRCA2 4859delA mutation. This mutation was found in an
the BRCA1 R1835X). The BRCA2 4265delCT mutation, previ- ovarian cancer patient who was investigated for a breast mass at
ously described for a Swedish family (BIC), was detected in 2 the breast clinic of the Philippine General Hospital (found to be
unrelated incident breast cancer cases from the Philippine General non-malignant). The 4859delA mutation was found for a woman
Hospital. In addition, the 4265delCT mutation was found in a with a previous diagnosis of childhood-onset non-Hodgkin’s lym-
prevalent breast cancer patient included in the haplotype study. phoma (PH0356) and for an ovarian cancer patient of Filipino
Table V depicts the genotypes associated with the BRCA2 origin ascertained in Canada as part of an Ontario-wide study of
4265delCT, 4859delA and the BRCA1 5454delC mutations. incident ovarian cancer cases (FOTS10285)27 (Table II). Micro-
Two Filipino 4265delCT patients of Malay ancestry shared a satellite analysis showed that 3 of the 7 4859delA cases had a
haplotype for markers both proximal and distal of BRCA2. One common haplotype for markers both proximal to and distal to
Filipino 4265delCT case (PH0093) of Hispanic/Chinese origin BRCA2. Two of the remaining 4859delA cases originated from the
exhibited a haplotype that can be related to the haplotype of the 2 Eastern Visayan island of Leyte and shared a haplotype that could
other 4265delCT mutation carriers, if recombination occurred be- be related to the haplotype of the other 3 4859delA carriers by
tween BRCA2 and D13S1701, although an independent origin of inferring a recombination between BRCA2 and D13S1701. Figure
 BRCA1 AND BRCA2 MUTATONS IN THE PHILIPPINES 601

FIGURE 2 – Cumulative incidence of breast cancer in female first-

degree relatives (BRCA1/2 mutation carriers vs. non-carriers). The
percentage of female relatives affected with breast cancer is displayed
by age (graph shown is a 1-minus cumulative survival Kaplan-Meier
curve). All subjects included in the analysis were censored at time of
death, cancer diagnosis or age at date of interview with proband. We
observed 6 cases of breast cancer among the 63 female relatives of
case patients with a BRCA1 or BRCA2 mutation, and 31 cases of breast
cancer among the 979 female relatives of case patients who did not
carry a mutation.

3 displays a map of the Philippines with the origin of patients with

the BRCA2 4265delCT and the 4859delA mutations.
Three BRCA1 mutations were identified in this study (Table II).
The Q1538X mutation has not previously been described, and the
R1835X mutation has been reported on multiple occasions for FIGURE 3 – Map of Philippines and origin of patients with founder
families of German and Dutch origin (BIC). The BRCA1 5454delC BRCA2 mutations. The BRCA2 4859delA mutation was identified in 7
mutation was found for a Filipino breast cancer patient residing in unrelated cancer patients, including the 4 patients with incident breast
Hawaii (personal communication). Family members of this woman cancers described in our study. This mutation may originate from the
were offered genetic counseling and predictive testing at the Phil- Eastern Visayan island of Leyte, as reported by 3 subjects. The BRCA2
ippine General Hospital in Manila (PH0561). Genotypes at 3 4265delCT mutation was identified in 3 patients, including the 2
patients with incident breast cancers described in our study. Two
BRCA1 intragenic markers were examined for 2 unrelated patients women originated from the Southern Tagalog region of the main
with the BRCA1 5454delC mutation. These data support an ances- island of Luzon. One patient with the BRCA2 4265delCT mutation
tral link (Table V). originated from another Eastern Visayan island, Samar. The map was
modified from CNN web-site [[Link]

DISCUSSION
The high prevalence of BRCA2 mutations in the Philippines is
This is the first report of the prevalence of BRCA1 and BRCA2 attributable in large part to 2 founder BRCA2 mutations in exon 11
mutations in a large unselected series of patients with breast cancer (4265delCT and 4859delA). These accounted for 6 of the 12
in Asia. Fifteen of 294 women (5.1%) diagnosed with invasive families with BRCA2 mutations. The BRCA2 3827delGT mutation
breast cancer at the Philippine General Hospital in Manila carried was detected in 2 study subjects. Only 1 of the other 6 BRCA2
a pathogenic BRCA1 or BRCA2 mutation. The overall prevalence mutations has been reported previously, i.e., the BRCA2 2042insA
of BRCA1 and BRCA2 mutations may be an underestimate because mutation was reported in 4 German families (BIC). Previous
the sensitivity of the mutation detection techniques used in our studies have estimated the prevalence of BRCA1 mutations in
study is not complete, estimated at less than 80%. Our study unselected East Asian patients with breast cancer, but none of
population represents a small sample of the total number of women these studies have included BRCA2. Therefore, the contribution of
diagnosed with breast cancer at the PGH in Manila, and it may be BRCA2 to breast cancer in Asian populations may not be appre-
that the true prevalence of BRCA mutations in the Philippines is ciated.
higher than 5.1%. Nonetheless, the prevalence of BRCA2 muta- Recently, 2 founder BRCA2 mutations were described for Jap-
tions is among the highest reported to date: most studies of anese breast cancer families (S2834X and 5802del4). Ikeda et al.28
unselected breast cancer cases typically report a greater proportion found that 21 of 101 site-specific breast cancer families (i.e., 2 or
of mutations in BRCA1 than in BRCA2. 3 cases of breast cancer and no ovarian cancer) were attributable to
In the Philippines, mutations in BRCA1 and BRCA2 do not a mutation in BRCA2 (20.7%), and 8 families were attributable to
contribute equally to breast cancer. Only 3 BRCA1 mutations were BRCA1 mutations (7.9%). None of the Japanese BRCA2 families
identified in our study population (1%). Two families with the included cases of ovarian cancer and 76% had only 2 cases of
5454delC mutation appear to share a common haplotype, indicat- breast cancer. Only 1 of the BRCA2-associated breast cancer
ing a probable ancestral link. patients from the Philippines had a relative with ovarian cancer
602 DE LEON MATSUDA ET AL.

(Fig. 1). These studies support the hypothesis that the clinical unaffected carrier relatives. The standard treatment of breast can-
expression of the BRCA2 mutation may differ between Asian and cer is currently mastectomy, but options for women with a BRCA
Caucasian carriers. mutation also include bilateral mastectomy, oophorectomy, tubal
The frequency of the founder Philippine mutations in the gen- ligation and tamoxifen. Female relatives may be offered predictive
eral population is largely unknown. Two of our controls were testing, and if found to be positive, prophylactic measures such as
found to carry 1 of the 2 BRCA2 founder mutations, resulting in a mastectomy, oophorectomy and tamoxifen should be discussed.
frequency of 0.6%. This prevalence is very high (1 in 170), and Breast screening is not common in the Philippines and compliance
identical to the frequency of the 999del5 in the Icelandic popula- is poor.30 Few women perform breast self-examination or undergo
tion10 but is based on small numbers, and the majority of these regular clinical breast examination. As a result, the majority of
women were selected because of the presence of a breast mass. It women present with advanced disease; in our study 63% had
will be important to confirm the mutation prevalence in a larger tumors larger than 5 cm (T3 or T4) and 67% had lymph node
population. Most of our controls were selected for benign breast involvement.
disease, but this is not associated with a BRCA1 or BRCA2 muta- Among the 294 unselected Manila cases, there were 8 cases that
tion. Also, the 2 controls with mutations did not have strong family belonged to families with hereditary breast-ovarian cancer syn-
histories of breast cancer. drome. Among these, 3 mutations were detected. Therefore, if full
There are several possible reasons for the high prevalence of screening were to be performed for the cases that appeared to be
BRCA2 mutations observed. It is possible that women with a hereditary, at a cost of 4,500 dollars per screen, we would have
family history of breast cancer were more likely to seek evaluation invested 36,000 dollars and would have found 3 mutations (13,500
for their breast complaint. It is also possible that the higher dollars per mutation). If we were to perform PTT for BRCA1 and
proportion of BRCA2 mutations in breast cancer cases is due to a BRCA2 for the 246 cases diagnosed at or below age 50 (at 200
greater penetrance of BRCA2 than of BRCA1 in the Philippines. dollars per case), we would have invested 49,200 dollars and
This could be the case if the modifying genetic or environmental would have found 12 mutations (4,100 dollars per mutation). If we
factors are different for the 2 genes. Finally, there may be a were to screen only for the 2 BRCA2 and 1 BRCA1 founder
selective advantage to carriers of specific BRCA2 alleles, in terms mutations (at 50 dollars per case) on all 236 cases at or below age
of reproductive fitness. Healy et al.29 reported that a specific 50, we would invest 11,800 dollars to identify 7 mutations (1,686
BRCA2 variant (N372H) was associated with increased reproduc- dollars per mutation). In comparison, in North America, compre-
tive fitness in males in the United Kingdom and an increased breast hensive genetic testing is usually offered if there is a 10 –20%
cancer risk in females. probability of a mutation being detected (22,500 – 45,000 dollars
per mutation identified). Therefore, in countries such as the
Our control group was comprised mainly of women under Philippines where founder mutations are present, it is more
investigation for a breast complaint (mass, discharge or pain). It is economical to screen all cases for founder mutations than to
possible that in some cases, a family history of breast cancer may perform a complete 2-gene screen on the subset of familial
lead women in the Philippines to seek medical attention and this cases. Mutation analysis for a limited set of founder mutations
may explain in part why a similar proportion of cases and controls requires much less time, resources and labor than complete
had a first-degree relative affected with breast cancer (11% and sequencing, resulting in a significant reduction in cost per
9%, respectively). Consequently, we examined the risk to first- mutation detected, and a greater number of mutations will be
degree relatives using data from the Manila Cancer Registry. We found. It is reasonable to offer women diagnosed before age 50
observed a 3-fold increased risk of breast cancer in the female genetic testing for the founder mutations by multiplex PCR
first-degree relatives of cases compared with the general popula- assay. We describe a small number of founder mutations, but
tion (Table IV). others will undoubtedly follow. Women with a strong family
When comparing carrier cases to non-carrier cases, we observed history may benefit from additional testing.
that 7 of the 15 cases with a germline mutation had a first or These data provide us with the opportunity to use genetic
second-degree relative with breast cancer or ovarian cancer compared testing for breast cancer prevention and to individualize treat-
with 51 of 264 non-carriers (odds ratio 3.65; p ⫽ .019). The cumu- ment in the Philippines. The proportion of breast cancer cases
lative incidence of breast cancer among the female first-degree rela- attributable to BRCA mutations, in particular germline BRCA2
tives of mutation-positive cases was 24.3% to age 50 and was much mutations, is higher than for North America. However, it is
greater than the incidence among female first-degree relatives of important to recognize that the social and medical context
mutation-negative cases (⬍4%; p⫽ 7.2 ⫻ 10⫺5). In fact, the cumu- surrounding genetic testing is different in the Philippines than
lative risk estimate for relatives of carriers of BRCA mutations in the in North America, due to different cultural values, beliefs about
Philippines exceeded the risk for relatives of Jewish breast cancer the origin of cancer, public awareness and access to care.
cases with BRCA mutations in North America, which was obtained Because of the lack of an organized breast cancer screening
using identical methods20 (Fig. 2). Therefore, the penetrance of program and limited access to tamoxifen, and because of the
BRCA2 mutations is not reduced in the Philippines. difficulties with maintaining adequate patient follow-up, pro-
Our results have implications for the care of women with breast phylactic oophorectomy and mastectomy should be explored as
cancer and their families in the Philippines, and in other develop- options for primary prevention of breast cancer to carriers of
ing countries. There are several important differences between our BRCA mutations in the Philippines, as well as in other devel-
patient population and the majority of women who are diagnosed oping countries.
in North America. The women with breast cancer in our study
were referred to a government-sponsored charity hospital for the
ACKNOWLEDGEMENTS
investigation of a breast mass. These patients have limited re-
sources and their treatment options are restricted. Currently, sur- We thank the women and their families, without whose coop-
gery is provided without cost, but radiotherapy, chemotherapy and eration this study would not have been possible. We thank Dr. P.
tamoxifen are available only to those with sufficient means. Also, Holowaty, A.R. Manansala, R.D. Nituda, J. Tanedo, J. Aubé and P.
it is difficult for patients from outlying regions to attend the de los Rios for their contribution to this study. We are grateful to
hospital; in fact, several women traveled for days to consult with S. Donlon for contributing information on her breast cancer patient
specialists at the Philippine General Hospital. In consequence, the from Hawaii and referring this BRCA1 family to the Philippine
oncology team attempts to minimize the number of patient visits. General Hospital for predictive testing and genetic counseling. We
Patient follow-up is difficult and only half of these women have thank Dr. D.M. Parkin for reviewing the article and the Interna-
private telephones. Nevertheless, it is valuable to offer genetic tional Agency for Research in Cancer in Lyon for allowing us to
testing to newly diagnosed cases of breast cancer in the Philippines publish figures. AL’s doctoral studies are supported by the Cana-
for the purposes of clinical management and as a means to identify dian Institutes of Health Research (formerly MRC).
 BRCA1 AND BRCA2 MUTATONS IN THE PHILIPPINES 603
REFERENCES

1. Parkin DM, Whelan SL, Ferlay J, et al., eds. Cancer Incidence in Five and other sequence variants in a population-based sample of Austra-
Continents. IARC Scientific Publications No. 143. Lyon: IARC, 1997. lian women with breast cancer. Br J Cancer 1999;79:34 –9.
2. Anton-Culver H, Cohen PF, Gildea ME, et al. Characteristics of 19. Thorlacius S, Struewing JP, Hartge P, et al. Population-based study of
BRCA1 mutations in a population-based case series of breast and risk of breast cancer in carriers of BRCA2 mutation. Lancet 1998;352,
ovarian cancer. Eur J Cancer 2000;36:1200 – 8. 1337–9.
3. Langston AA, Malone KE, Thompson JD, et al. BRCA1 mutations in 20. Warner E, Foulkes W, Goodwin P, et al. Prevalence and penetrance of
a population-based sample of young women with breast cancer. BRCA1 and BRCA2 gene mutations in unselected ashkenazi jewish
N Engl J Med 1996;334:137– 42. women with breast cancer. J Natl Cancer Inst 1999;91:1241–7.
4. Malone KE, Daling JR, Neal C, et al. Frequency of BRCA1/BRCA2 21. Worsham MJ, Nathanson SD, Christopherson P, et al. A new BRCA1
mutations in a population-based sample of young breast carcinoma mutation in a Filipino woman with a family history of breast and
cases. Cancer 2000;88:1393– 402. ovarian cancer. Diagnostic Molec Pathol 1998;7:164 –167.
5. Malone KE, Daling JR, Thompson JD, et al. BRCA1 mutations and 22. Laudico AV, Esteban DB, Reyes LM, et al. Philippine Cancer Facts and
breast cancer in the general population: analyses in women before age Estimates, Manila, Philippines: Philippine Cancer Society, Inc., 1998.
35 years and in women before age 45 years with first-degree family 23. Fodde R, Losekoot M. Mutation detection by denaturing gradient gel
history. JAMA 1998;279:922–9. electrophoresis (DGGE). Hum Mutat 1994;3:83–94.
6. Newman B, Mu H, Butler LM, et al. Frequency of breast cancer 24. Neuhausen SL, Godwin AK, Gershoni-Baruch R, et al. Haplotype and
attributable to BRCA1 in a population-based series of American phenotype analysis of 9 recurrent BRCA2 mutations in 111 families:
women. JAMA 1998;279:915–1. results of an international study. Am J Hum Genet 1998;62:1381– 8.
7. Ellis D, Greenman J, Hodgson S, et al. Low prevalence of BRCA1 25. Neuhausen SL, Mazoyer S, Friedman L, et al. Haplotype and pheno-
mutations in early onset breast cancer without family history. J Med type analysis of 6 recurrent BRCA1 mutations in 61 families: results
Genet 2000;37:792– 4. of an international study. Am J Hum Genet 1996;58:271– 80.
8. Peto J, Collins N, Barfoot R, et al. Prevalence of BRCA1 and BRCA2 26. Breslow NE, Day NE, eds. Volume II. The Design and Analysis of
gene mutations in patients with early-onset breast cancer. J Natl Cohort Studies. Lyon: IARC Scientific Publication, 1987.
Cancer Inst 1999;91:943–9. 27. Risch HA, McLaughlin JR, Cole DEC, et al. Prevalence and penetrance
9. Johannesdottir G, Gudmundsson J, Bergthorsson JT, et al. High of germline BRCA1 and BRCA2 mutations in a population series of 649
prevalence of the 999del5 mutation in Icelandic breast and ovarian women with ovarian cancer. Am J Hum Genet 2001;68:700–10.
cancer patients. Cancer Res 1996;56:3663–5. 28. Ikeda N, Miyoshi Y, Yoneda K, et al. Frequency of BRCA1 and
10. Thorlacius S, Sigurdsson S, Bjarnadottir H, et al. Study of a single BRCA2 germline mutations in Japanese breast cancer families. Int J
BRCA2 mutation with high carrier frequency in a small population Cancer 2001;91:83– 8.
Am J Hum Genet 1997;60:1079 – 84. 29. Healey CS, Dunning AM, Dawn Teare M, et al. A common variant in
11. Diez Gilbert O, Machuca Cordano I, Angel Navarro M. Character- BRCA2 is associated with both breast cancer risk and prenatal via-
ization of the BRCA1 gene and its significance in hereditary breast bility. Nat Genet 2000;26:362– 4.
cancer. Medicina Clinica 1996;107:623–7. 30. Parkin DM, Pisani P, Gibson LJ, et al. Screening for cancer of the
12. Garcia-Patino E, Gomendio B, Provencio M, et al. Germ-line BRCA1 breast in the Philippines IARC Biennial Report, 1998/1999. Lyon:
mutations in women with sporadic breast cancer: clinical correlations. IARC, 2000. 129 –30.
J Clin Oncol 1998;16:115–20.
13. Papelard H, de Bock GH, van Eijk R, et al. Prevalence of BRCA1 in
a hospital-based population of Dutch breast cancer patients. Br J
Cancer 2000;83:719 –24. APPENDIX I—ELECTRONIC-DATABASE INFORMATION
14. Van Der Looij M, Szabo C, Besznyak I, et al. Prevalence of founder
BRCA1 and BRCA2 mutations among breast and ovarian cancer Accession numbers and URLs for data in this article are as
patients in Hungary. Int J Cancer 2000;86:737– 40. follows: Online Mendelian Inheritance in Man (OMIM), http://
15. Backe J, Hofferbert S, Skawran B, et al. Frequency of BRCA1 [Link]/Omim/, for inherited breast cancer type 1
mutation 5382insC in German breast cancer patients. Gynecol Oncol
1999;72:402– 6. and ovarian cancer [MIM 113705] and for inherited breast cancer
16. Syrjakoski K, Vahteristo P, Eerola H, et al. Population-based study of type 2 [MIM 600185]. Human Genome Database (HGD), http://
BRCA1 and BRCA2 mutations in 1035 unselected Finnish breast [Link]/, for DNA sequence primers of 13q and 17q loci.
cancer patients. J Natl Cancer Inst 2000;92:1529 –31. Breast Cancer Information Core Database (BIC), [Link]
17. Hopper JL, Southley MC, Dite GS, et al. Population-based estimate of [Link]/Intramural_research/Lab_transfer/ Bic/, for information
the average age-specific cumulative risk of breast cancer for a defined
set of protein-truncating mutations in BRCA1 and BRCA2. Cancer on sequence alterations in BRCA1 and BRCA2. PubMed (Medline
Epidemiol Biomarkers Prevention 1999;8:741–7. online), [Link] for literature
18. Southey MC, Tesoriero AA, Andersen CR, et al. BRCA1 mutations searches on population-specific studies of BRCA1 and BRCA2.