ORIGINAL RESEARCH

published: 28 January 2022

doi: 10.3389/fonc.2021.743231

Novel Insights From the Germline

Landscape of Breast Cancer in Brazil
Daniel Barbalho 1*, Renata Sandoval 2, Erika Santos 2, Janina Pisani 2, Carla Quirino 2,
Bernardo Garicochea 2,3, Benedito Rossi 2 and Maria Isabel Achatz 2
1 Department of Breast Surgery, Hospital Sirio-Libanês, São Paulo, Brazil, 2 Department of Oncogenetics, Hospital

Sirio-Libanês, São Paulo, Brazil, 3 Centro Paulista de Oncologia, Oncoclinicas, São Paulo, Brazil

Introduction: Breast cancer patients with germline pathogenic variants may benefit from
risk-reducing surgeries, intensive screening, and targeted cancer therapies. There is a
paucity of data regarding prevalence and distribution of germline pathogenic variants in
the Brazilian population. Our primary endpoint was the description of prevalence and
distribution of germline pathogenic variants among breast cancer patients who underwent
next-generation sequencing (NGS) panel testing. Secondary endpoint was the
Edited by:
Daniela Turchetti, assessment of predictive factors of a positive test.
University of Bologna, Italy
Methods: We analyzed NGS results, personal, and family history data from a
Reviewed by:
Maria Del Pilar Estevez Diz,
prospectively collected cohort of breast cancer patients from August 2013 to May
Universidade de São Paulo, Brazil 2019. Exact logistic regression was used to perform multivariable analysis.
Natalija Dedić Plavetić,
University Hospital Centre Zagreb,
Results: Of 370 breast cancer patients, we found 59 pathogenic variants in 57 (15%)
Croatia patients. Pathogenic variants were identified in BRCA1 (24%), ATM (14%), BRCA2 (10%),
*Correspondence: TP53 (8%), PALB2 (8%), CHEK2 (7%), CDH1 (3%), RAD51C (3%), MITF (2%), PMS2
Daniel Barbalho
(2%), RAD51D (2%), and TERT (2%). Monoallelic MUTYH pathogenic variants were found
[email protected]
in 15%. After multivariable analysis, age of diagnosis (OR 0.89, 95% CI: 0.81–0.95, for
Specialty section: each year increase), triple-negative subtype (OR 17.2, 95% CI: 3.74–114.72), and number
This article was submitted to
of breast cancers in the family (OR 2.46, 95% CI 1.57–4.03, for each additional case) were
Cancer Genetics,
a section of the journal associated with BRCA1 pathogenic variants. In the present study, a quarter of triple-
Frontiers in Oncology negative breast cancer patients harbored a germline pathogenic variant and two-thirds of
Received: 17 July 2021 those were BRCA1 carriers.
Accepted: 31 December 2021
Published: 28 January 2022 Conclusions: Prevalence and distribution of germline pathogenic variants in this Brazilian
Citation: sample of breast cancer patients are mostly similar to other populations. However, there is
Barbalho D, Sandoval R, Santos E, a trend to an overrepresentation of TP53 pathogenic variants that merits confirmation in
Pisani J, Quirino C, Garicochea B,
Rossi B and Achatz MI (2022) Novel further studies. Early-onset breast cancer patients should be offered genetic counseling,
Insights From the Germline Landscape particularly those with triple-negative subtype.
of Breast Cancer in Brazil.
Front. Oncol. 11:743231. Keywords: breast neoplasms, high-throughput nucleotide sequencing, Brazil, genetic predictive testing, genetic
doi: 10.3389/fonc.2021.743231 predisposition to breast cancer

Frontiers in Oncology | www.frontiersin.org 1 January 2022 | Volume 11 | Article 743231

Barbalho et al. Germline Breast Cancer in Brazil

INTRODUCTION (bilateral cases count as 2 and personal history was excluded),

number of male breast cancer, cancer of the ovary, pancreas,
Breast cancer affects approximately 66,000 women and accounts prostate, melanoma, sarcoma, adrenocortical, central nervous
for more than 17,000 deaths annually in Brazil (1). system (CNS), leukemia, gastric, colon, endometrium, thyroid,
Approximately 10% of breast cancer patients carry a germline and kidney.
pathogenic variant that may indicate screening strategies or New assessment of variants segregation within families was
preventive recommendations (2). Targeted therapy options not possible. However, we already had information on the
may be indicated as further treatment. segregation of 20 pathogenic variants. In the absence of
After the identification of the BRCA1 gene in 1994 by Dr. segregation information, either maternal or paternal family
Mary-Claire King (3), DNA sequencing techniques and history was collected based on the following criteria, in order
bioinformatics have evolved significantly (4) and rendered of priority: number of additional breast cancer cases, youngest
germline testing accessible to an increasingly wider population. additional breast cancer case, degree of relationship to proband,
Since then, other high penetrance genes such as TP53 and PALB2 and the presence of ovarian, pancreatic, sarcoma, or central
have also been described as breast cancer susceptibility genes, as nervous system cancers. There were no ties beyond this point.
well as moderate penetrance genes such as ATM and CHEK2 (5). Primary endpoint was the description of prevalence and
Recommendations for BRCA1 and BRCA2 range from risk- distribution of germline pathogenic variants among breast
reducing mastectomies (6) to intensive screening with breast cancer patients that had NGS testing. Secondary endpoint was
magnetic resonance imaging (MRI) (7) and specific therapies the assessment of predictive clinical factors of a positive test.
such as platinum agents (8) and PARP inhibitors (9, 10). There is Likely Pathogenic Variants were regarded as Pathogenic and
an ongoing effort to understand the magnitude and the Variants of Unknown Significance as Not Pathogenic.
modifying factors of risk conferred by each gene and to which Continuous data were not normally distributed and are presented
extent we could generalize what we learned from high penetrance as median and interquartile range. Since most of family history data
genes to moderate penetrance counterparts (11). had a median of zero, data are presented as categorical (at least 1 case
As there is a paucity of data describing the germline landscape of each cancer) in order to better disclose clinical significance, but
of breast cancer patients in the Brazilian population, we aimed to they were treated as continuous variables for the purpose of statistical
describe the prevalence and distribution of germline pathogenic inference. Categorical data are presented as percentiles. Univariate
variants among breast cancer patients in a tertiary oncology analysis was performed with Mann–Whitney’s or Fisher’s exact test
hospital in Brazil. for continuous and categorical data, respectively. For the purpose of
multivariable analysis, exact logistic regression was performed for
each gene in a forward selection manner from the most significant
one, until there was no significant covariate left behind. Since there
METHODS were 13 models, Bonferroni correction was applied to account for
multiple testing. Therefore, a significant p-value was set at 0.0038.
This is a cross-sectional study from a prospectively collected There was less than 8% of missing data in histology and breast cancer
database from the Oncogenetics Unit at Hospital Sı́rio-Libanês, subtype only. Missingness was not related to any other variable.
a tertiary oncology hospital based in São Paulo and Brası́lia, Brazil. Hence, it was assumed to be missing completely at random and dealt
From August 2013 to May 2019, 2,116 subjects were included in with the worst-case scenario. All analyses were performed using the
the registry, and 867 had a personal history of breast cancer. software Stata 17.
Among these, 386 had non-NGS testing, and 97 did not have a Patients prospectively signed an informed consent form to have
sample collected. Eligible subjects were breast cancer patients who their data and family history collected and used for research
received genetic counseling in this time frame and to whom next- purposes. Institutional Review Board approved data collection as
generation-sequencing (NGS) cancer panel was performed no new medical intervention would be pursued and confidentiality
(Invitae™ 83 or 84 Multi-Cancer Panel) (12). Patients were would be preserved. Data were de-identified for the purpose of
referred based on their physician’s assessment of risk factors for statistical analysis and protected from re-identification.
hereditary cancer. The indication of germline testing followed
NCCN criteria, but 10% of our sample were offered testing without
a formal criterion, mainly because of second malignancies or the RESULTS
presence of multiple breast cancer cases above age 50 in the family.
Tests were paid out-of-pocket and reimbursed by insurance In total, 384 charts were electronically reviewed. One subject was
companies, whenever applicable. Exclusion criteria were inability excluded for not having an NGS panel test, five for not having a
to retrieve data and absence of family history records. personal history of breast cancer and eight for substantial
All medical charts were electronically reviewed, and the following missing data. Among the remaining 370 subjects, 59
data were collected: NGS panel results, age of personal breast cancer pathogenic variants were identified in 57 (15%) subjects. Two
diagnosis, gender, Ashkenazi ethnicity, personal history of bilateral subjects had 2 pathogenic variants concomitantly, one being the
breast cancer, histology, and immunohistochemistry (IHC) subtype. combination of TP53 and ATM pathogenic variants and the
Family history was collected up to third-degree relatives including other TP53 and monoallelic MUTYH. In the span of six years, 62
personal history, and comprehended number of breast cancer cases variants were reclassified, most of them from unknown

Frontiers in Oncology | www.frontiersin.org 2 January 2022 | Volume 11 | Article 743231

Barbalho et al. Germline Breast Cancer in Brazil

significance to benign, but there was one CHEK2 intronic variant prevalence in the pathogenic group. A quarter of triple-negative
that was reclassified as likely pathogenic. As of January 2020, patients harbored a germline pathogenic variant, and two-thirds
there were 178 (48%) variants of unknown significance of those were BRCA1 carriers. Eighty percent of BRCA1 carriers
identified. The distribution of pathogenic variants can be had a triple-negative cancer (Tables 1 and 4).
found in Figure 1. Median number of breast cancer cases in the family was
Ninety percent of our population had at least 1 criterion similar between groups, but there was 1.5 extra case in BRCA1
according to the National Comprehensive Cancer Network carriers. Prevalence of ovarian and pancreatic cancers was
(NCCN) guidelines, version 1.2020. There were only 2 male doubled in the pathogenic group (Table 2).
subjects, and only 4 with Ashkenazi ethnicity. None of them Sarcoma was a rare event, but at least four times more
carried a germline pathogenic variant. Median age at breast frequent in the pathogenic group. Numerically, there were
cancer diagnosis was 46 in patients with no pathogenic more cases of melanoma, endometrium, and kidney cancers, as
variants. Median age was 5 years younger in the pathogenic well as less cases of leukemia and gastric cancers in the
group and 10 years younger in BRCA1 carriers. Bilateral cancers pathogenic group. Prostate, CNS, colon, and thyroid cancers
were twice more frequent in the pathogenic group (Table 1). were well balanced between groups (Table 2).
Groups were similar according to histology and subtype, Multivariable analysis was significant only for BRCA1 after
except for the triple-negative subtype, which was doubled in correction for multiple testing. Younger age, triple-negative
subtype, and number of breast cancer cases in the family were
highly correlated with the presence of a BRCA1 pathogenic
variant (Table 3).
Complete description of germline pathogenic variants and
cases can be found in Table 4.

DISCUSSION
This is a highly selected convenience sample from a tertiary
oncology hospital in Brazil. Median age at breast cancer
diagnosis was 45, while previous studies have found it to be 54
in a Brazilian sample (13), and 62 in SEER registry (14).
Notwithstanding this selection, having a BRCA1 pathogenic
variant was significantly associated with age at diagnosis,
further lowering median age to 34, with 75th percentile at 42.
Germline pathogenic variants prevalence at 15% is in
FIGURE 1 | Distribution of germline pathogenic variants.
accordance with previous studies. The true prevalence lies at

TABLE 1 | Personal characteristics of the study population.

Total (%) n = 370 (100%) Not Pathogenic (%) n = 313 (85%) Pathogenic (%) n = 57 (15%) Univariate Analysis p-value

Female 368 (99) 311 (99) 57 (100) ns

Ashkenazi Ethnicity 4 (01) 4 (01) 0 (00) ns

Median age at diagnosis (IQR) 45 (39–52) 46 (40–53) 41 (35–48) 0.0038

Bilateral Cancers (07) 18 (06) 8 (14) 0.042

At least 1 NCCN criterion 332 (90) 277 (88) 55 (96) ns

Ductal Carcinoma In Situ 42 (11) 37 (12) 5 (09) ns

Ductal Carcinoma 253 (68) 211 (67) 42 (74) ns
Lobular Carcinoma 36 (10) 31 (10) 5 (09) ns
Histology Others 22 (06) 20 (06) 2 (04) ns
Histology Missing Data 17 (05) 14 (04) 3 (05) ns

HR positive Her2 negative 236 (64) 203 (65) 33 (58) ns

HR positive Her2 positive 27 (07) 23 (07) 04 (07) ns
HR negative Her2 positive 30 (08) 27 (09) 3 (05) ns
HR negative Her2 negative 63 (17) 46 (15) 17 (30) 0.011
Subtype Missing Data* 27 (07) 23 (07) 4 (07) ns

IQR, interquartile range; HR, Hormone Receptors; Her2, Human Epidermal growth factor Receptor 2; ns, not significant; NCCN, National Comprehensive Cancer Network.
*Data do not sum up to 100% due to bilateral cancers.

Frontiers in Oncology | www.frontiersin.org 3 January 2022 | Volume 11 | Article 743231

Barbalho et al. Germline Breast Cancer in Brazil

TABLE 2 | Cancer family history.

Total (%) n = 370 (100%) Not Pathogenic (%) Pathogenic (%) Univariate
n = 313 (85%) n = 57 (15%) Analysis p-value

Median Number of Breast Cancer Cases in the 1 (0–2) 1 (0–2) 1 (0–2) 0.0402
Family up to Third Degree, excluding proband (IQR)
At least 1 Cancer in the Family up to Third Degree,
including proband:
Male Breast 4 (01) 4 (01) 0 (00) ns
Ovary 35 (09) 26 (08) 9 (16) ns
Pancreas 34 (09) 25 (08) 9 (16) ns
Prostate 79 (21) 67 (21) 12 (21) ns
Melanoma 26 (07) 20 (06) 6 (11) ns
Sarcoma 12 (03) 7 (02) 5 (09) 0.0105
Adrenocortical 0 (00) 0 (00) 0 (00) NA
Central Nervous System 28 (08) 24 (08) 4 (07) ns
Leukemia 29 (08) 27 (09) 2 (04) ns
Gastric 37 (10) 34 (11) 3 (05) ns
Colon 93 (25) 78 (25) 15 (26) ns
Endometrium 8 (02) 6 (02) 2 (04) ns
Thyroid 30 (08) 24 (08) 6 (11) ns
Kidney 14 (04) 10 (03) 4 (07) ns

IQR, interquartile range; ns, not significant; NA, not applicable.

TABLE 3 | Multivariable analysis.

Variable BRCA1 (%) Not BRCA1 (%) Odds Ratio p-value

Median Age at Diagnosis (IQR) 34.5 (32–42) 45 (39.5–52) OR 0.89 (0.81–0.95) 0.0005
Triple-Negative Subtype 11/14 (79) 75/356 (21) OR 17.20 (3.74–114.72) <0.0001
Median Number of Breast Cancer Cases in 2.5 (1–3) 1 (0–2) OR 2.46 (1.57–4.03) 0.0001
the Family up to Third Degree, excluding proband (IQR)

IQR, interquartile range; OR, Odds Ratio.

approximately 10%, according to the largest series published to and in Asian countries was approximately 1.0%. The largest
date (2), but that ranged from 6% in a study from the Mayo series from the USA reports a TP53 prevalence of 0.17% (2),
Clinic (15), to 34% from Stanford University (12). The higher whereas in Asian countries it was 1.9% in China (17), 1.0% in
prevalence from Stanford can be partially explained by the South Korea (18), and 1.5% in Taiwan (21). This was not
recruitment period in which testing criteria were more replicated in other Latin American countries other than Brazil.
stringent, whereas the study from Mayo already included There were no reports of the TP53 R337H in breast cancer cohort
patients when NGS technology was commonly available. studies from Argentina, Colombia, Guatemala, Mexico, or Peru
In addition, only 47.9% of Mayo’s sample had at least 1 NCCN (16, 22).
criterion, and 29.9% of identified pathogenic variants came from Of note, all TP53 variants identified in this study were the
subjects without any criterion. That led the authors to propose R337H, described by Achatz et al. (23) as associated to Li-
access to germline testing for all breast cancer patients diagnosed Fraumeni syndrome, albeit with a later onset of disease. Even
below age 65. In this study, having at least 1 NCCN criterion was though we did not find any adrenocortical tumor in our sample,
not associated to the presence of a pathogenic variant. We cannot this variant has been linked to this cancer in the pediatric
reach the same conclusion solely based on our sample, because population of Brazil (24). In addition, this variant was
the majority (90%) of our subjects had at least 1 criterion. identified at a surprisingly high rate (0.21%) among 35,000
Distribution of variants was likewise in accordance with previous newborns from an unselected population in the Southeast
studies, being roughly a third to a half in BRCA1 and BRCA2, and region of Brazil (25).
the remaining among ATM, CHEK2, PALB2, and TP53. The frequency of triple-negative cancers at 17% is also in
We performed an unplanned exploratory analysis comparing accordance with previous studies (13, 14). BRCA1 carriers had
our results to formerly published ones from different countries 80% of triple-negative cancers, an association long recognized in
(Table 5). In this study, the frequency of TP53 variants at 1.3% the literature (26). Triple-negative subtype was an important
was 6.2 times higher than the pooled results from previous positive predictive factor, as a quarter of triple-negative cancers
reports, p = 0.002, although we acknowledge our limited was linked to a pathogenic variant, and two-thirds of these
absolute number. While the prevalence of TP53 variants variants were in BRCA1, an important finding that has clinical
among all available data is 0.2%, the frequency in our sample implications in the therapeutic and prophylactic settings.

Frontiers in Oncology | www.frontiersin.org 4 January 2022 | Volume 11 | Article 743231

Barbalho et al. Germline Breast Cancer in Brazil

TABLE 4 | Description of germline pathogenic variants.

Patient Germline Pathogenic Variant Age at Bilateral Triple Cancer Family History up to Third Degree (Age) Second
Diagnosis Cancer Negative Malignancies of
Proband (Age)

01 ATM c.1339C>T (p.Arg447Ter) 39 No No Breast (40) and Colon (58).

02 ATM c.2999del (p.Asn1000Thrfs*2) 35 No No Breast (65) and Breast (72).
03 ATM c.3802delG (p.Val1268*) 41 No No Breast (54) and Breast (56).
04 ATM c.3802delG (p.Val1268*) 33 No No None.
05 ATM c.4741dupA (p.Ile1581Asnfs*5) 48, 67 Yes Yes Breast (64), Breast (80), Leukemia (80), and Kidney (81).
06 ATM c.4906C>T (p.Gln1636Ter) 42, 46 Yes No Breast (40), Gastric, Colon (73), and Kidney.
07 ATM c.67C>T (p.Arg23*) 45 No No Breast (54), Breast, Colon (80), Kidney (40), and Kidney (71).
08 ATM Partial Deletion (Exon 27-29) 66 No No Breast (30), Pancreas, Prostate (85), Sarcoma, and Thyroid
(15).
09 BRCA1 c.1071dup 34 No Yes Breast (34), Bilateral Breast (35, 35), and Breast (36).
(p.Leu358Thrfs*8)
10 BRCA1 c.1687C>T (p.Gln563*) 35, 35 Yes Yes Ovary (58).
11 BRCA1 c.188T>A (p.Leu63Ter) 46 No Yes Pancreas (64) and Gastric (72).
12 BRCA1 c.2176_2177delCT 39 No No Breast (42), Breast (69), Prostate (68), and Prostate (75).
(p.Leu726Serfs)
13 BRCA1 c.3770_3771delAG 41, 43 Yes Yes Breast (50), Breast (80), and Melanoma (43).
(p.Glu1257Glyfs*9)
14 BRCA1 c.4165_4166delAG 32 No Yes Breast (55).
(p.Ser1389*)
15 BRCA1 c.441+2T>A (Splice donor) 42 No Yes Breast (38), Breast (38), Breast (60), and Ovary (48).
16 BRCA1 c.5074+2T>C (Splice 29, 39 Yes Yes Breast (45), Breast (48), and Breast (55).
donor).
17 BRCA1 c.5266dupC 28 No – Breast (36), Breast, Breast, and Pancreas (70). Colon (70).
(p.Gln1756Profs)
18 BRCA1 c.5266dupC 62 No No Breast (28), Breast, Prostate, and Prostate. Lymphoma (62).
(p.Gln1756Profs)
19 BRCA1 c.5266dupC 48, 55 Yes Yes Breast (40), Bilateral Breast (43, 54), Breast (64), Ovary (80), Ovary (52).
(p.Gln1756Profs) Pancreas (42), and Prostate (67).
20 BRCA1 c.5266dupC 32 No Yes Breast (70).
(p.Gln1756Profs)
21 BRCA1 c.5554_5555delAC 32 No Yes Breast (40) and Colon (55).
(p.Thr1852Leufs*27)
22 BRCA1 c.798_799del 32 No Yes Bilateral Breast (38, 40), Breast (45), Ovary (40), and Ovary
(p.Ser267Lysfs*19) (50).
23 BRCA2 c.1138del 28 No No Breast (47) and Breast (70).
(p.Ser380Valfs*19)
24 BRCA2 c.6034del 32 No No Breast (78), Colon (70), and Thyroid (60).
(p.Ser2012Profs*28)
25 BRCA2 c.7007G>A (p.Arg2336His) 41 No No Breast (60) and Breast (65).
26 BRCA2 c.8009C>T (p.Ser2670Leu) 57 No No Gastric (59).
27 BRCA2 c.8878C>T (p.Gln2960Ter) 29 No No Ovary (35), Pancreas (73), Melanoma (58), and Melanoma
(63).
28 BRCA2 c.9097dupA 53 No No Breast (20), Breast (44), Breast (58), and Breast (70).
(p.Thr3033Asnfs*11)
29 CDH1 c.1763_1764del 47 No – Breast (48) and Colon (70). Melanoma (41).
p.Val588Glufs*2
30 CDH1 c.471dup (p.Ile158Tyrfs*10) 47 No No Bilateral Breast (32, 47), Colon (63), Colon (64), and Colon
(69).
31 CHEK2 c.1100del C (p.T367Mfs*15) 48 No No Prostate (43).
32 CHEK2 c.1459C>T (p.Gln487*) 41 No No Prostate (70), Thyroid (42), Thyroid (55), and Thyroid (60).
33 CHEK2 c.846+1G>C (Splice donor) 42 No No Breast, Breast, Breast, Pancreas, and Kidney (75). Parotid (49) and
Kidney (62).
34 CHEK2 c.846+4_846+7del (Intronic) 40 No No Leukemia (35).
35 MITF c.952G>A (p.Glu318Lys) 57 No No Prostate (49), Prostate (70), and Colon (78).
36 Monoallelic MUTYH c.1147delC 41 No No Sarcoma (70).
(p.Ala385Profs)
37 Monoallelic MUTYH c.1187G>A 35 No No Breast (60), Ovary (64), Prostate (72), Prostate (75),
(p.Gly396Asp) Melanoma (56), Central Nervous System (70), and Colon (48).
38 Monoallelic MUTYH c.1187G>A 53 No No Breast, Breast, Ovary (50), and Melanoma.
(p.Gly396Asp)

(Continued)

Frontiers in Oncology | www.frontiersin.org 5 January 2022 | Volume 11 | Article 743231

Barbalho et al. Germline Breast Cancer in Brazil

TABLE 4 | Continued

Patient Germline Pathogenic Variant Age at Bilateral Triple Cancer Family History up to Third Degree (Age) Second
Diagnosis Cancer Negative Malignancies of
Proband (Age)

39 Monoallelic MUTYH c.1187G>A 53 No No Pancreas (73), Prostate (65), and Prostate (80). Thyroid (32).
(p.Gly396Asp)
40 Monoallelic MUTYH c.1187G>A 40 No No Breast and Pancreas.
(p.Gly396Asp)
41 Monoallelic MUTYH c.1187G>A 47 No Yes Breast (73) and Endometrium (67).
(p.Gly396Asp)
42 Monoallelic MUTYH 34 No No Breast (42).
c.1437_1439delGGA (p.Glu480del)
43 Monoallelic MUTYH c.536A>G 40 No No Breast (47), Ovary, and Colon.
(p.Tyr179Cys)
44 Monoallelic MUTYH Deletion (Exons 58 No No Breast (50), Breast (70), Ovary (89), and Sarcoma (06).
4-16)
45 PALB2 c.1240C>T (p.Arg414*) 54, 67 Yes Yes Breast and Breast. Thyroid (54).
46 PALB2 c.1671_1674delTATT 50 No – Breast and Breast.
(p.Ile558Lysfs)
47 PALB2 c.1671_1674delTATT 38 No No Breast (67) and Colon (65).
(p.Ile558Lysfs)
48 PALB2 c.355delC (p.Gln119Lysfs) 38 No No Breast (50). Thyroid (36).
49 PALB2 Deletion (Exons 7-10) 41 No No Breast (41), Breast (60), and Prostate (75).
50 PMS2 c.903G>T (p.Lys301Asn) 60 No No Colon (63) and
Endometrium (69).
51 RAD51C c.709C>T (p.Arg237*) 51, 51 Yes Yes Prostate (87).
52 RAD51C Deletion (Exons 6-9) 43 No Yes Melanoma (65) and Central Nervous System (70). Gastrointestinal
Stromal Tumor
(45).
53 RAD51D c.694C>T (p.Arg232*) 34 No No Breast (45), Pancreas (69), and Colon (60).
54 TERT c.336dupC (p.Glu113Argfs) 38 No Yes
06 TP53 c.1010G>A (p.Arg337His) 42, 46 Yes No Breast (40), Gastric, Colon (73), and Kidney.
37 TP53 c.1010G>A (p.Arg337His) 35 No No Breast (60), Ovary (64), Prostate (72), Prostate (75),
Melanoma (56), Central Nervous System (70), and Colon (48).
55 TP53 c.1010G>A (p.Arg337His) 38 No No Breast (41), Breast (55), Prostate (75), Central Nervous Sarcoma (29).
System (15), and Colon (55).
56 TP53 c.1010G>A (p.Arg337His) 47 No – Breast (50), Breast (60), Central Nervous System (01), and Lung (56).
Central Nervous System (56).
57 TP53 c.1010G>A (p.Arg337His) 44 No No Breast (66) and Sarcoma (12).

As the assessment of variants segregation within families was among carriers of TP53 pathogenic variants (29). In our study,
not possible, family history was collected based on the criteria two out of five families with a TP53 variant had a sarcoma case,
described in the Methods section. One limitation of this study is one being the proband.
that tumors collected in family history could be sporadic, rather To our best knowledge, this is the largest series of breast cancer
than associated to the pathogenic variant found in the proband. patients in the Brazilian population in which all subjects had NGS
Nevertheless, family history is an easily accessible information in multigene panel testing. Palmero et al. described 229 BRCA1 and
clinical practice, and genetic testing of all relatives up to third BRCA2 variants identified in 28 centers across Brazil in subjects at
degree is rarely available in real life. high risk for hereditary breast or ovarian cancer, regardless of the
The association between pancreatic cancer and BRCA1 is well sequencing method (30). There are four novel variants described
established, with up to 10% of familial pancreatic cancer being in this article: BRCA1 c.1071dup, BRCA1 c.5554_5555delAC,
attributable to either a BRCA1 or BRCA2 variant (27). In our BRCA2 c.6034del, and BRCA2 c.8009C>T. Timoteo et al.
study, there were 4 cases of pancreatic cancers among 20 BRCA1 reported the distribution of pathogenic variants among 157
and BRCA2 carriers, 2.5 times the frequency in the group with no breast cancer patients, or at high risk for hereditary breast
pathogenic variants identified. cancer, in the state of Rio Grande do Norte. The overall
Pathogenic variants in CHEK2 have not been traditionally linked prevalence was 15%, with 11 variants in BRCA1 (07%), 5 in
to renal cell carcinoma, but they were the most prevalent germline BRCA2 (03%), 4 in ATM (03%), 1 in ATR (01%), 1 in CDH1
alteration (3.5%) in a study of 254 advanced renal cell carcinomas (01%), and 1 in MLH1 (01%) (31). Felix et al. analyzed 106
(28). In our study, one family out of 4 with CHEK2 variants subjects at high risk for hereditary breast cancer in the state of
presented two cases of renal cell carcinoma at ages 62 and 75. Bahia (32). BRCA1 was completely sequenced and specific variants
Sarcoma and TP53 is another well-established association in BRCA2, CHEK2, and TP53 were assessed. They found 9 variants
with a cumulative incidence of approximately 20% up to age 70 in BRCA1, and one R337H variant in TP53. Gomes et al. studied

Frontiers in Oncology | www.frontiersin.org 6 January 2022 | Volume 11 | Article 743231

Barbalho et al. Germline Breast Cancer in Brazil

TABLE 5 | Prevalence and istribution of germline pathogenic variants in breast cancer patients from different populations.

Brazil Latin-America† China South-Korea USADana-Farber USAMayo USAMyriad USAStanford† Italy† Taiwan
(%) (16) (%) (17) (%) (18) (%) (19) (%) (15) (%) (2) (%) (12) (%) (20) (%) (21) (%)

N 370 222 937 496 488 3907 35409 198 255 133
Patdogenic 57 (15) 31 (14) 215 (23) 79 (16) 52 (11) 246 (06) 3388 (10) 68 (34) 68 (27) 28 (21)
APC 0 (00) 0 (00) 0 (00) 0 (00) 0 (00) – 11 (00) 0 (00) 0 (00) 0 (00)
ATM 8 (02) 1 (00) 6 (01) 0 (00) 4 (01) 43 (01) 329 (01) 2 (01) 3 (01) 1 (01)
BARD1 0 (00) – 5 (01) 0 (00) 0 (00) – 68 (00) – – 0 (00)
BLM 0 (00) 0 (00) – 0 (00) – – – 1 (01) 0 (00) –
BRCA1 14 (04) 10 (04) 82 (09) 31 (06) 18 (04) 51 (01) 814 (02) 35 (18) 32 (13) 9 (07)
BRCA2 6 (02) 14 (06) 81 (09) 30 (06) 12 (02) 56 (01) 828 (02) 24 (12) 26 (10) 11 (08)
BRIP1 0 (00) 0 (00) 3 (00) 1 (00) 4 (01) – 110 (00) 0 (00) 2 (01) 1 (01)
CDH1 2 (01) 0 (00) 2 (00) 8 (02) 0 (00) 6 (00) 23 (00) 1 (01) 0 (00) 0 (00)
CDKN2A 0 (00) 1 (00) 0 (00) 0 (00) 0 (00) – 32 (00) 2 (01) 0 (00) –
CHEK2 3 (01) 0 (00) 6 (01) 2 (00) 10 (02) 67 (02) 397 (01) – 0 (00) 0 (00)
EPCAM 0 (00) 0 (00) – 0 (00) 0 (00) – 4 (00) 0 (00) 0 (00) 0 (00)
MITF 1 (00) – – – – – – – 0 (00) –
MLH1 0 (00) 0 (00) 1 (00) 2 (00) 0 (00) – 22 (00) 1 (01) 0 (00) 0 (00)
MSH2 0 (00) 1 (00) 3 (00) 1 (00) 0 (00) – 37 (00) 0 (00) 0 (00) 1 (01)
MSH6 0 (00) 1 (00) 0 (00) 0 (00) 1 (00) – 73 (00) 0 (00) 1 (00) 0 (00)
Monoallelic 9 (02) 3 (01) 8 (01) 1 (00) 9 (02) – – 5 (03) – 1 (01)
MUTYH
Biallelic 0 (00) 0 (00) 0 (00) 0 (00) 0 (00) – 7 (00) 0 (00) – 0 (00)
MUTYH
NBN 0 (00) 0 (00) 0 (00) 3 (01) 1 (00) – 59 (00) 2 (01) 0 (00) 0 (00)
NF1 0 (00) 0 (00) 0 (00) 0 (00) – 1 (00) – – 0 (00) –
PALB2 5 (01) 2 (01) 11 (01) 0 (00) 1 (00) 15 (00) 316 (01) 0 (00) 6 (02) –
PMS2 1 (00) 0 (00) 2 (00) 0 (00) 1 (00) – 101 (00) 0 (00) 0 (00) 0 (00)
PTEN 0 (00) 0 (00) 0 (00) 0 (00) 1 (00) 1 (00) 17 (00) 0 (00) 0 (00) 0 (00)
RAD50 0 (00) – 2 (00) 0 (00) – – – – – 2 (01)
RAD51C 2 (01) 0 (00) 0 (00) 0 (00) 1 (00) – 53 (00) 0 (00) 0 (00) 1 (01)
RAD51D 1 (00) 0 (00) 0 (00) 0 (00) 1 (00) – 19 (00) – 1 (00) 0 (00)
RECQL4 0 (00) 0 (00) – – – – – – 1 (00) –
SMAD4 0 (00) 0 (00) – 0 (00) 0 (00) – 3 (00) 0 (00) 0 (00) 0 (00)
STK11 0 (00) 0 (00) 0 (00) 0 (00) 0 (00) – 4 (00) 0 (00) 0 (00) 0 (00)
TERT 1 (00) – 0 (00) – – – – – 0 (00) –
TP53 5 (01)* 0 (00) 18 (02) 5 (01) 0 (00) 6 (00) 61 (00) 0 (00) 0 (00) 2 (01)
TSC2 0 (00) 0 (00) – – – – – – 1 (00) –
WRN 0 (00) 1 (00) 0 (00) – – – – – 0 –

*p = 0.002. †Sample also included subjects with Hereditary Breast and Ovarian Cancer Syndrome without a Breast Cancer Diagnosis.

126 patients in the state of Rio de Janeiro, with either breast or is warranted to clarify the true prevalence and meaning
ovarian cancer, that had at least 1 NCCN criterion and no of specific variants from Brazil, particularly the TP53
pathogenic variants in BRCA1 or BRCA2 (33). They found one R337H variant. Early-onset breast cancer patients should be
variant in ATM, two in CHEK2, one in PALB2, and one in TP53. offered genetic counseling, particularly those with triple-
Furthermore, this study aimed to assess predictive factors of a negative subtype.
positive test in addition to describing variants. Even though we
were not able to detect any novel predictive factor, this article
sums to the body of evidence regarding the genetic germline
landscape of Brazilian breast cancer patients. It is also a call for DATA AVAILABILITY STATEMENT
more research on the true prevalence of TP53 variants in The datasets presented in this study can be found in online
unselected subjects, and on the elucidation of clinical repositories. The names of the repository/repositories and
implications of specific variants from Brazil. a c c e s s i o n n u m b e r ( s ) ca n b e f o u n d i n t h e a r t i c l e /
supplementary material.

CONCLUSIONS
ETHICS STATEMENT
Prevalence and distribution of germline pathogenic variants
in this Brazilian sample of breast cancer patients are mostly The studies involving human participants were reviewed and
similar to other populations. However, there is a trend to an approved by Comitê de É tica em Pesquisa em Seres Humanos do
overrepresentation of TP53 pathogenic variants. Future research Hospital Sı́r io-Libanê s . Written informed consent for

Frontiers in Oncology | www.frontiersin.org 7 January 2022 | Volume 11 | Article 743231

Barbalho et al. Germline Breast Cancer in Brazil

participation was not required for this study in accordance with All authors contributed to the article and approved the
the national legislation and the institutional requirements. submitted version.

AUTHOR CONTRIBUTIONS SUPPLEMENTARY MATERIAL

DB is the first author. MA is the senior author. ES, JP, and CQ The Supplementary Material for this article can be found online at:
contributed equally to data access and management. RS, BG, and https://www.frontiersin.org/articles/10.3389/fonc.2021.743231/
BR contributed equally to data interpretation and text revision. full#supplementary-material

16. Oliver J, Quezada Urban R, Franco Corté s CA, Dı́az Velá squez CE,
REFERENCES Montealegre Paez AL, Pacheco-Orozco RA, et al. Latin American Study of
1. De Oliveira Santos M. Estimativa/2020 – Incidência De Câncer No Brasil. Rev Hereditary Breast and Ovarian Cancer LACAM: A Genomic Epidemiology
Bras Cancerol (2020) 66(1):1. doi: 10.32635/2176-9745.rbc.2020v66n1.927 Approach. Front Oncol (2019) 9:1429. doi: 10.3389/fonc.2019.01429
2. Buys S, Sandbach JF, Gammon A, Patel G, Kidd J, Brown K, et al. A Study of Over 17. Li JY, Jing R, Wei H, Wang M, Xiaowei Q, Liu H, et al. Germline Mutations in
35,000 Women With Breast Cancer Tested With a 25-Gene Panel of Hereditary 40 Cancer Susceptibility Genes Among Chinese Patients With High
Cancer Genes. Cancer (2017) 123(10):1721–30. doi: 10.1002/cncr.30498 Hereditary Risk Breast Cancer. Int J Cancer (2019) 144(2):281–9.
3. King MC. “The Race” to Clone BRCA1. Sci (80-) (2014) 343(6178):1462–5. doi: 10.1002/ijc.31601
doi: 10.1126/science.1251900 18. Shin H-C, Lee H-B, Yoo T-K, Lee ES, Kim RN, Park B, et al. Detection of
4. Shendure J, Balasubramanian S, Church GM, Gilbert W, Rogers J, Schloss JA, Germline Mutations in Breast Cancer Patients With Clinical Features of
et al. DNA Sequencing at 40: Past, Present and Future. Nature (2017) 550 Hereditary Cancer Syndrome Using a Multi-Gene Panel Test. Cancer Res
(7676):345–53. doi: 10.1038/nature24286 Treat (2020) 52(3):697–713. doi: 10.4143/crt.2019.559
5. Couch FJ, Shimelis H, Hu C, Hart SN, Polley EC, Na J, et al. Associations 19. Tung N, Lin NU, Kidd J, Allen BA, Singh N, Wenstrup RJ, et al. Frequency of
Between Cancer Predisposition Testing Panel Genes and Breast Cancer. Germline Mutations in 25 Cancer Susceptibility Genes in a Sequential Series
JAMA Oncol (2017) 3(9):1190–6. doi: 10.1001/jamaoncol.2017.0424 of Patients With Breast Cancer. J Clin Oncol (2016) 34(13):1460–8.
6. Hunt KK, Euhus DM, Boughey JC, Chagpar AB, Feldman SM, Hansen NM, doi: 10.1200/JCO.2015.65.0747
et al. Society of Surgical Oncology Breast Disease Working Group Statement 20. Tedaldi G, Tebaldi M, Zampiga V, Danesi R, Arcangeli V, Ravegnani M, et al.
on Prophylactic (Risk-Reducing) Mastectomy. Ann Surg Oncol (2017) Multiple-Gene Panel Analysis in a Case Series of 255 Women With
(2):375–97. doi: 10.1245/s10434-016-5688-z Hereditary Breast and Ovarian Cancer. Oncotarget (2017) 8(29):16791.
7. Mann RM, Kuhl CK, Moy L. Contrast-Enhanced MRI for Breast Cancer doi: 10.18632/oncotarget.16791
Screening. J Magn Reson Imaging (2019) 50(2):377–90. doi: 10.1002/ 21. Lin PH, Kuo WH, Huang AC, Lu YS, Lin CH, Kuo SH, et al. Multiple Gene
jmri.26654 Sequencing for Risk Assessment in Patients With Early-Onset or Familial
8. Tutt A, Tovey H, Cheang MCU, Kernaghan S, Kilburn L, Gazinska P, et al. Breast Cancer. Oncotarget (2016) 7(7):8310–20. doi: 10.18632/
Carboplatin in BRCA1/2-Mutated and Triple-Negative Breast Cancer oncotarget.7027
BRCAness Subgroups: The TNT Trial. Nat Med (2018) (5):628–37. 22. Gallardo-Alvarado LN, Tusié -Luna MT, Tussié -Luna MI, Dı́az-Chá vez J,
doi: 10.1038/s41591-018-0009-7 Segura YX, Bargallo-Rocha E, et al. Prevalence of Germline Mutations in
9. Robson ME, Tung N, Conte P, Im SA, Senkus E, Xu B, et al. OlympiAD Final the TP53 Gene in Patients With Early-Onset Breast Cancer in the Mexican
Overall Survival and Tolerability Results: Olaparib Versus Chemotherapy Population. BMC Cancer (2019) 19(1):118. doi: 10.1186/s12885-019-
Treatment of Physician’s Choice in Patients With a Germline BRCA 5312-2
Mutation and HER2-Negative Metastatic Breast Cancer. Ann Oncol (2019) 23. Achatz MI, Zambetti GP. The Inherited P53 Mutation in the Brazilian
30(4):558–66. doi: 10.1093/annonc/mdz012 Population. Cold Spring Harb Perspect Med (2016) 6(12):a026195.
10. Litton JK, Rugo HS, Ettl J, Hurvitz SA, Gonçalves A, Lee KH, et al. Talazoparib doi: 10.1101/cshperspect.a026195
in Patients With Advanced Breast Cancer and a Germline BRCA Mutation. 24. Ferreira AM, Brondani VB, Helena VP, Charchar HLS, Zerbini MCN, Leite
N Engl J Med (2018) 379(8):753–63. doi: 10.1056/NEJMoa1802905 LAS, et al. Clinical Spectrum of Li-Fraumeni Syndrome/Li-Fraumeni-Like
11. Tung NM, Boughey JC, Pierce LJ, Robson ME, Bedrosian I, Dietz JR, et al. Syndrome in Brazilian Individuals With the TP53 P.R337H Mutation. J
Management of Hereditary Breast Cancer: American Society of Clinical Steroid Biochem Mol Biol (2019) 190:250–5. doi: 10.1016/j.jsbmb.2019.04.011
Oncology, American Society for Radiation Oncology, and Society of 25. Custó dio G, Parise GA, Filho N, Komechen H, Sabbaga CC, Rosati R, et al.
Surgical Oncology Guideline. J Clin Oncol (2020) 38(18):2080–106. Impact of Neonatal Screening and Surveillance for the TP53 R337H Mutation
doi: 10.1200/JCO.20.00299 on Early Detection of Childhood Adrenocortical Tumors. J Clin Oncol (2013)
12. Kurian AW, Hare EE, Mills MA, Kingham KE, McPherson L, Whittemore AS, 31(20):2619–26. doi: 10.1200/JCO.2012.46.3711
et al. Clinical Evaluation of a Multiple-Gene Sequencing Panel for Hereditary 26. Shimelis H, LaDuca H, Hu C, Hart SN, Na J, Thomas A, et al. Triple-Negative
Cancer Risk Assessment. J Clin Oncol (2014) 32(19):2001–9. doi: 10.1200/ Breast Cancer Risk Genes Identified by Multigene Hereditary Cancer Panel
JCO.2013.53.6607 Testing. J Natl Cancer Inst (2018) 110(8):855–62. doi: 10.1093/jnci/djy106
13. Simon SD, Bines J, Werutsky G, Nunes JS, Pacheco FC, Segalla JG, et al. 27. Pilarski R. The Role of BRCA Testing in Hereditary Pancreatic and Prostate
Characteristics and Prognosis of Stage I-III Breast Cancer Subtypes in Brazil: Cancer Families. Am Soc Clin Oncol Educ B (2019) 39:79–86. doi: 10.1200/
The AMAZONA Retrospective Cohort Study. Breast (2019) 44:113–9. edbk_238977
doi: 10.1016/j.breast.2019.01.008 28. Carlo MI, Mukherjee S, Mandelker D, Vijai J, Kemel Y, Zhang L, et al.
14. Surveillance, Epidemiology, and End Results (SEER). Program Populations Prevalence of Germline Mutations in Cancer Susceptibility Genes in Patients
(1969-2019). National Cancer Institute, DCCPS, Surveillance Research With Advanced Renal Cell Carcinoma. JAMA Oncol (2018) 4(9). doi: 10.1001/
Program, released February 2021. Available at: (www.seer.cancer.gov/ jamaoncol.2018.1986
popdata). 29. Mai PL, Best AF, Peters JA, DeCastro RM, Khincha PP, Loud JT, et al. Risks of
15. Yadav S, Hu C, Hart SN, Boddicker N, Polley EC, Na J, et al. Evaluation of First and Subsequent Cancers Among TP53 Mutation Carriers in the National
Germline Genetic Testing Criteria in a Hospital-Based Series of Women With Cancer Institute Li-Fraumeni Syndrome Cohort. Cancer (2016) 122
Breast Cancer. J Clin Oncol (2020) 38(13):1409–18. doi: 10.1200/jco.19.02190 (23):3673–81. doi: 10.1002/cncr.30248

Frontiers in Oncology | www.frontiersin.org 8 January 2022 | Volume 11 | Article 743231

Barbalho et al. Germline Breast Cancer in Brazil

30. Palmero EI, Carraro DM, Alemar B, Moreira MAM, Ribeiro-Dos-Santos Conflict of Interest: The authors declare that the research was conducted in the
 , Abe-Sandes K, et al. The Germline Mutational Landscape of BRCA1 absence of any commercial or financial relationships that could be construed as a
and BRCA2 in Brazil. Sci Rep (2018) 8(1):1–10. doi: 10.1038/s41598-018- potential conflict of interest.
27315-2
31. de Souza Timoteo AR, Gonçalves AÉ MM, Sales LAP, Albuquerque BM, de Publisher’s Note: All claims expressed in this article are solely those of the authors
Souza JES, de Moura PCP, et al. A Portrait of Germline Mutation in Brazilian and do not necessarily represent those of their affiliated organizations, or those of
at-Risk for Hereditary Breast Cancer. Breast Cancer Res Treat (2018) 172 the publisher, the editors and the reviewers. Any product that may be evaluated in
(3):637–46. doi: 10.1007/s10549-018-4938-0 this article, or claim that may be made by its manufacturer, is not guaranteed or
32. Felix GES, Abe-Sandes C, Machado-Lopes TMB, Bomfim TF, Guindalini RS, endorsed by the publisher.
Santos VC, et al. Germline Mutations in BRCA1, BRCA2, CHEK2 and TP53
in Patients at High-Risk for HBOC: Characterizing a Northeast Brazilian Copyright © 2022 Barbalho, Sandoval, Santos, Pisani, Quirino, Garicochea, Rossi and
Population. Hum Genome Var (2014) 1:14012. doi: 10.1038/hgv.2014.12 Achatz. This is an open-access article distributed under the terms of the Creative
33. Gomes R, Spinola P da S, Brant AC, Matta BP, Nascimento CM, de Aquino Commons Attribution License (CC BY). The use, distribution or reproduction in other
Paes SM, et al. Prevalence of Germline Variants in Consensus Moderate-to- forums is permitted, provided the original author(s) and the copyright owner(s) are
High-Risk Predisposition Genes to Hereditary Breast and Ovarian Cancer in credited and that the original publication in this journal is cited, in accordance with
BRCA1/2-Negative Brazilian Patients. Breast Cancer Res Treat (2021) 185 accepted academic practice. No use, distribution or reproduction is permitted which
(3):851–61. doi: 10.1007/s10549-020-05985-9 does not comply with these terms.

Frontiers in Oncology | www.frontiersin.org 9 January 2022 | Volume 11 | Article 743231