ORIGINAL ARTICLE

DOI: 10.29289/2594539420190000436

EVALUATION OF BONE METASTASIS OF BREAST

CANCER TO LONG OR SHORT BONES, ACCORDING
TO MOLECULAR SUBTYPES: RETROSPECTIVE STUDY
Avaliação da metástase óssea do câncer de mama em ossos longos ou
curtos, segundo os subtipos moleculares: estudo retrospectivo
Paulo Roberto de Andrade Figaro Caldeira1* , Carlos Augusto Real Martinez1 , José Roberto Fígaro Caldeira2

ABSTRACT
Bone is the most frequent site for breast cancer metastasis. Identifying the possible preference of bone metastasis, such as long or
short bones, according to molecular subtypes, could alter oncologists approach, paying special attention to these particular group
of patients reducing the side effects of the bone metastatic process, involving multidisciplinary team with orthopedists, minimizing
possible sequelae of this metastatic process. Detecting different metastatic sites to long or short bones, according to the molecular
subtypes and their possible correlation. Fifty-eight patients with only bone metastasis were chosen. The study material was obtained
from paraffin embedded primary tumors. Statistical analysis of the data was carried out. The luminal A, luminal B, hybrid luminal,
HER2 + and triple-negative / basal-like molecular subtypes were identified. The molecular subtypes compared to the age of bone
implants, the distribution of bone implants, and the disease free interval were not statistically significant.
KEYWORDS: Molecular biology; breast cancer; neoplasm metastasis.

RESUMO
Acometimento ósseo é o sítio mais comum de metástase do carcinoma de mama. A identificação de possível preferência conforme
os subtipos moleculares, na precocidade ou no acometimento de ossos longos ou chatos, poderia alterar a prática médica de
oncologistas, dirigindo especial atenção a esses grupos de pacientes e suas possíveis complicações, em atendimento multidisciplinar
com ortopedistas, minimizando possíveis sequelas desse processo metastático. Detectar a instalação dos diferentes sítios
metastáticos para ossos longos ou chatos (curtos), conforme os subtipos moleculares e sua possível correlação. Foram selecionados
58 casos de pacientes com câncer de mama que apresentaram exclusivamente metástases ósseas. O material de estudo foi obtido
dos tumores primários emblocados em parafina. Realizaram-se análises estatísticas dos dados. Foram identificados os subtipos
moleculares luminal A, luminal B, luminal híbrido, HER2+ e triplo-negativo/basal like. Os subtipos moleculares comparados com a
idade de implantes ósseos, a distribuição de implantes ósseos e o intervalo livre de doença não mostraram significância estatística.
PALAVRAS-CHAVE: Biologia molecular; neoplasias da mama; metástase neoplásica.

Study carried out at Hospital Amaral Carvalho – Jaú (SP), Brazil.

1
Department of Surgical Clinic, Universidade São Francisco – Bragança Paulista (SP), Brazil.
2
Department of Mastology, Hospital Amaral Carvalho – Jaú (SP), Brazil.
*Corresponding author: [email protected]
Conflict of interests: nothing to declare.
Received on: 10/11/2018. Accepted on: 12/27/2018

32 Mastology, 2019;29(1):32-36
 Relationship between bone metastasis, molecular subtypes and breast cancer

INTRODUCTION was reported by taking luminal A from breast carcinoma as a

Bones represent the most common site of distant metastasis reference. Triple-negative breast cancer demonstrated large
of breast carcinoma. Bones from different parts of the skel- tropism for lung, while the non-luminal subtype human epi-
eton, especially short (flat) bones, are often compromised by dermal growth factor type 2 (HER2) was associated with
metastatic dissemination in women with breast cancer. It is high rate of liver metastasis. All subtypes were associated
not well understood why the initial mechanism of metastatic with low risk of bone only location. Briefly, this study added
implants has a greater preference for bones. Among the short information to understand the complexity of breast cancer
(flat) bones, the sequence of impairment for sternum, ribs, ver- and its clinical manifestations. It also proposes categorization
tebrae and pelvis is observed. Short (flat) bones are affected between different subgroups based on the immunohistochemi-
before long ones1. cal resources, as it could predict the preferential anatomical
Bone metastasis is strongly associated with positive estrogen site of the first distant metastasis, as well as specific prog-
receptor/negative progesterone receptor in tumors. Significant dif- nosis. It is therefore tempting to hypothesize some practical
ference in tumors with estrogen receptor expression, between implication in terms of “adapted” management, i.e., surveil-
high and low grade with bone metastasis, suggests that differ- lance protocols and/or therapeutic strategies that need to be
ent panels of molecular markers could be used to predict bone verified by clinical trials 4.
metastasis in these two groups of tumors2. Differences in the biological characteristics of breast can-
The average time to diagnosis of only breast cancer metasta- cer can be explained by differences in the pattern of changes
sis from the last follow-up or death was 55.2 months. Only bone between genes that act on carcinogenesis. Several studies have
metastasis have been reported to occur in 17-37% of patients with been conducted to determine the value of genetic changes as
distant metastasis. Metastasis to the confined distance to the prognostic markers for these patients. The molecular prog-
skeleton presents a more favorable prognosis than other types of nostic markers used in clinical practice are: estrogen receptor
distant metastasis or multiple metastasis to bones and viscera. (ER), progesterone receptor (PR) and tyrosine kinase receptor
Other investigators reported that the median survival of patients (ERBB2 / HER2). The presence or absence of these proteins
with bone metastasis alone was 24-54 months. The favorable fea- is commonly detected using immunohistochemistry analy-
ture of the primary tumor accounts for the modest prognosis of sis. Thus, three main molecular classes were established:
women with only bone metastasis3. positive hormone receptor tumors, HER2 positive tumors
There is great evidence on the differences in dissemina- and negative tumors for all the markers used. These classes
tion among the biological subtypes of breast cancer. A study have been integrated into diagnosis and treatment and help
performed to analyze the metastatic pattern according to to stratify the risk of recurrence, especially in lymph node
the biological subtype explores the corresponding progno- negative patients 5 .
sis. Biological subtype was defined by immunohistochem- Involvement of axillary lymph nodes is considered the
istry according to the criterion of St. Gallen, 2013, Swiss most informative prognostic factor. In practice, patients with
city where annual meetings of oncologists occur, in which four or more positive lymph nodes are considered a subgroup
consensus of prognoses and treatments are constructed, as of unfavorable prognosis 6 . In the year 2000, Perou et al. pub-
adapted in Table 1. Association between biological subtypes lished a work that became a reference to classify breast can-
and the distant and different locations were analyzed. Result cers in molecular subtypes, according to the gene expression
pattern: luminal A, luminal B, superexpressor HER2, basaloid
and normal-like7.
Table 1. Immunophenotypic profile to approximate molecular
According to Barros and Leite, these tumor subgroups
classification in breast carcinoma.
present varying patterns of behavior regarding the expres-
Molecular subtype Profile of biomarkers
sion of genes, the rate of tumor growth, as well as prognosis
Luminal A RE+ and/or RP+; HER2-; Ki-67<14% and sensitivity to treatment. According to these authors, the
Luminal B RE+ and/or RP+; HER2-; Ki-67≥14% luminal subtype A corresponds to 30-40% of the cases; lumi-
Luminal hybrid RE+ and/or RP+; HER2+ nal B, 20 to 30%; and HER2 and basaloid, from 15 to 20% of
HER2+ RE-; RP-; HER2+ the sample 8 .
Triple negative RE-; RP-; HER2-
Basal like RE-; RP-; HER2-; CK 5/6+ and/or EGFG+
OBJECTIVE
Source: Hammond et al.10; Cheang et al.11; Wolff et al.12; Wludarski and
Bacchi13; Cheang et al.14; Bhargava et al.15. To detect the installation of different metastatic sites for long or
ER: estrogen receptor; PR: progesterone receptor; HER2: human epidermal flat (short) bones in breast cancer, according to the molecular
growth factor type 2; Ki-67: protein encoded by the MKI67 gene; EGFG:
epidermal growth factor gene. subtypes and their possible correlation.

Mastology, 2019;29(1):32-36 33
 Caldeira PRAF, Martinez CAR, Caldeira JRF

METHOD The histological classification of the tumors evaluated in this

This study is a historical cohort, in which 58 cases of invasive study is organized in Table 3.
breast carcinoma, exclusively affected by bone metastasis, attended
by the Department of Mastology of Hospital Amaral Carvalho, About molecular subtypes and the detection of
Jaú, São Paulo, were retrospectively selected between January implants in long bones, short (flat) bones or both
2000 and January 2012. The present study was approved by the As shown in Table 4, of the total of 58 cases, the tendency to be
Research Ethics Committee of the Hospital Amaral Carvalho and implanted in flat bones in the luminal molecular subtypes was
the Plataforma Brasil, under No. 1.546.684, dated May 16th, 2016. evidenced, totaling 24 cases. In long bones, three cases were
Patients with breast cancer exclusively presenting bone obtained, and in both types (long and flat), eight cases, total-
metastasis from breast carcinoma; who underwent immunohis- ing 35 cases.
tochemistry and adjuvant chemotherapy, according to the pro-
tocol of the Clinical Oncology Department of Hospital Amaral
Carvalho; with adjuvant radiotherapy treatment, if indicated; DISCUSSION
with hormone therapy with tamoxifen or aromatase inhibitor, The investigation of exclusively bone metastasis becomes diffi-
if necessary, according to the hormonal (positive) receptor sta- cult, since the metastasis are usually implanted simultaneously,
tus, were accepted for the present study. in multiple sites4.
Patients with distant metastasis reaching bones, viscera (lung The breast tumor samples from these 58 patients were clas-
and liver), central nervous system and skin (synchronic metas- sified according to type and histological degree.
tasis to different sites) were excluded from the present study. Of this total, 51 cases were classified as ductal carcinomas,
whose histological grade ranged from 1 to 3, being 1 well differ-
Identification of metastatic sites entiated and 3 undifferentiated. The majority found was histo-
Metastatic sites were identified by imaging bone scintigraphy, logical grade 2, that is, moderately differentiated. The other forms
radiography, computed tomography and nuclear magnetic reso- found were mucinous carcinoma (one case), lobular infiltrating
nance, when indicated. (four cases) and apocrine carcinoma (one case), and one case
Regarding the metastatic sites in the bones, these were sub- without histological classification.
divided into three groups: long bones, short (flat) bones and both.
The long bones considered were: femur, tibia, fibula, humerus,
Table 3. Histological types and respective classification of
radius, ulna and clavicle. And among short or flat bones: bones
histological grade.
of the skull, spine, sternum, ribs and pelvis.
Histological types Number of cases
The routine immunohistochemical analysis was done with the
collaboration of Dr. Francisco Carlos Quevedo and Dr. Francisco Infiltrating ductal carcinoma G3 15

Alves Moraes Neto, Department of Pathology, Hospital Amaral Infiltrating ductal carcinoma G2 35
Carvalho, Jaú. Infiltrating ductal carcinoma G1 1
To facilitate the analysis of this work, and in view of tumor Mucinous carcinoma 1
biological behavior, the molecular subtypes were grouped into Infiltrating lobular carcinoma 4
four groups in Table 2. Apocrine carcinoma 1
No classification for histological rating 1
Total cases 58
RESULTS
The results are described in the form of tables and graphics. The sta-
tistical results are indicated with their corresponding p-value; and
Table 4. Distribution of bone metastasis according to molecular
the tests are named when necessary.
subtypes.
Flat
Bones/Molecular subtypes Long Both Total
Table 2. Groups of molecular subtypes. (short)
Grouping Subtypes Luminal A 1 16 3 20
Subtype 1 Luminals A and B Luminal B 2 8 5 15
Subtype 2 Group HER2+ Hybrid 1 5 - 6
Subtype 3 Hybrid luminal group Triple-negative 1 5 5 11
Subtype 4 Triple-negative and basal-like group HER2 - 4 2 6
HER2: human epidermal growth factor type 2. HER2: human epidermal growth factor type 2.

34 Mastology, 2019;29(1):32-36
 Relationship between bone metastasis, molecular subtypes and breast cancer

These numbers are in agreement with the literature data, New studies, especially using a larger sample, are necessary
since ductal carcinomas represent 80% of the breast tumors, to affirm or not some relation of what was studied here.
and the lobular tumors, approximately 10%. The other forms rep-
resent 1% of breast cancers, in their respective classifications9.
In this sample of 58 cases, the immunohistochemical analy- CONCLUSIONS
sis revealed 35 cases classified as luminal molecular subtypes A Due to the heterogeneity of its clinical and histopathological
and B; 6 cases, HER2+ subtype; 6 cases, hybrid luminal subtype; presentation and the difficulty of selecting cases of metastatic
and 11 cases, triple-negative/basal like. These numbers were cor- breast cancer exclusively for bone, the present study met the pro-
roborated by Barros and Leite in a recent review article8. posed objectives and was able to conclude:
The analysis of the correlation between the molecular sub- • Bone metastasis were found in long bones, flat bones or both,
types of breast carcinomas (luminal A and B, luminal hybrid, tri- depending on the molecular subtypes of breast carcinoma
ple-negative/basal like, HER2+) and implants for flat bones, long and their possible correlations. Of the 58 cases analyzed,
bones or both (Table 4) observed 58 metastatic cases, 24 cases for 38 were implanted in flat bones, thus distributed: 24 in the
flat bones in luminal molecular subtypes; 3 cases in long bones luminal subtypes, 5 in the hybrid, 5 in the triple negative and
and 8 cases in long and flat bones, totaling 35 cases. 4 in the HER2. As to the implant in long bones, 5 cases were
It is known that 60 to 70% of mammary tumors are of the lumi- identified, thus distributed: 3 in the luminal subtypes, 1 in the
nal molecular subtype A and B8. In the sample, we identified 35 cases hybrid and 1 in the triple-negative/basal like. Regarding the
of luminal bone implants, a prevalence of 60.34%, considered high occurrence of both types of bones, 15 metastatic implants
in comparison to other molecular subtypes. These data are cor- were found: 8 in luminal subtypes, 5 in triple-negative/basal-
roborated by the finding in the literature, according to Wei et al. 2. like and 2 in HER2;
It should be noted that 24 of the 35 metastatic luminal cases • The molecular subtypes of breast tumors classified as luminal
were only for flat bones, that is, approximately 70% of the cases. (A and B), triple-negative/basal-like, HER2 group and hybrid
These findings, in general, can contribute to the clinical prac- luminal were identified by immunohistochemical reaction.
tice of oncologists, especially mastologists, in light of the fact that It has also been observed that luminal molecular subtypes
luminal subtypes have a preference for bone implants, with 60% form the majority of bone metastasis.
corresponding to flat bones.
Thus, clinical practice is recommended for care in the first Finally, these data also indicate the need for molecular-
months of follow-up after surgery, especially in cases of luminal level research using these common molecular subtypes of
subtypes, for the request of bone scintigraphy in the search for breast cancer in the search for possible tumor markers for
possible bone metastasis. bone metastasis.

REFERENCES
1. Piato S, Piato JRM. Doenças da mama. Rio de Janeiro: Revinter; 5. Robison JE, Perreard L, Bernard PS. State of the science:
2006. 312p molecular classifications of breast cancer for clinical
2. Wei B, Wang J, Bourne P, Yang Q, Hicks D, Bu H, et al. Bone diagnostics. Clin Biochem. 2004;37(7):572-8. https://doi.
metastasis is strongly associated with estrogen receptor-positive/ org/10.1016/j.clinbiochem.2004.05.002
progesterone receptor-negative breast carcinomas. Hum Pathol. 6. Faneyte IF, Peterse JL, Van Tinteren H, Pronk C, Bontenbal
2008;39(12):1809-15. https://doi.org/10.1016/j.humpath.2008.05.010 M, Beex LV, et al. Predicting early failure after adjuvant
3. Ahn SG, Lee HM, Cho SH, Lee SA, Hwang SH, Jeong J, et al. chemotherapy in high-risk breast cancer patients with
Prognostic factors for patients with bone-only metastasis in extensive lymph node involvement. Clin Cancer Res.
breast cancer. Yonsei Med J. 2013;54(5):1168-77. https://doi. 2004;10(13):4457-63. https://doi.org/10.1158/1078-0432.CCR-
org/10.3349/ymj.2013.54.5.1168 03-0054
4. Gerratana L, Fanotto V, Bonotto M, Bolzonello S, Minisini M, 7. Perou CM, Sorlie T, Eisen MB, Van de Rijn M, Jeffrey
Fasola G, et al. Pattern of metastasis and outcome in patients SS, Ress CA, et al. Molecular portraits of human breast
with breast cancer. Clin Exp Metastasis. 2015;32(2):125-33. tumors. Nature. 2000;406(6797):747-52. https://doi.
https://doi.org/10.1007/s10585-015-9697-2 org/10.1038/35021093

Mastology, 2019;29(1):32-36 35
 Caldeira PRAF, Martinez CAR, Caldeira JRF

8. Barros ACSD de, Leite KRM. Classificação molecular dos 12. Wolff AC, Hammond MEH, Schwartz JN, Hagerty KL, Allred
carcinomas de mama: uma visão contemporânea. Rev Bras DC, Cote RJ, et al. American Society of Clinical Oncology/
Mastologia. 2015;25(4):146-55. https://dx.doi.org/10.5327/ College of American Pathologists guideline recommendations
Z201500040006RBM for human epidermal growth factor receptor 2 testing in breast
9. Arpino G, Bardou VV, Clark GM, Elledge RM. Infiltrating cancer. J Clin Oncol. 2007;25(1):118-45. https://doi.org/10.1200/
lobular carcinoma of the breast: tumor characteristics and JCO.2006.09.2775
clinical outcome. Breast Cancer Res. 2004;6(3):R149-56. 13. Wludarski SCL, Bacchi CE. High concordance of SP3
https://dx.doi.org/10.1186%2Fbcr767 rabbit monoclonal antibody with FISH to evaluate
10. Hammond ME, Hayes DF, Dowsett M, Allred DC, Hagerty HER2 in breast carcinoma. Appl Immunohistochem Mol
KL, Badve S, et al. American Society of Clinical Oncology/ Morphol. 2008;16(5):466-70. http://dx.doi.org/10.1097/
College of American Pathologists guideline recommendations PAI.0b013e318162625c
for immunohistochemical testing of estrogen and 14. Cheang MCU, Chia SK, Voduc D, Gao D, Leung S, Snider J, et
progesterone receptors in breast cancer. Arch Pathol Lab Med. al. Ki67 index, HER2 status, and prognosis of patients with
2010;134(7);e48-72. luminal B breast cancer. J Natl Cancer Inst. 2009;101(10):736-
11. Cheang MC, Treaba DO, Speers CH, Olivotto IA, Bajdik CD, 50. https://doi.org/10.1093/jnci/djp082
Chia SK, et al. Immunohistochemical detection using the new 15. Bhargava RB, Beriwal S, Dabbs DJ, Ozbek U, Soran A, Johnson
rabbit monoclonal antibody SP1 of estrogen receptor in breast RR, et al. Immunohistochemical surrogate markers of breast
cancer is superior to mouse monoclonal antibody 1D5 in cancer molecular classes predicts response to neoadjuvant
predicting survival. J Clin Oncol. 2006;24(36):5637-44. https:// chemotherapy. Cancer. 2010;116(6):1431-9. https://doi.
doi.org/10.1200/JCO.2005.05.4155 org/10.1002/cncr.24876

36 Mastology, 2019;29(1):32-36