Seminars in Cancer Biology 31 (2015) 111–118

Contents lists available at ScienceDirect

Seminars in Cancer Biology journal homepage: www.elsevier.com/locate/semcancer

Review

Evolving paradigms in multifocal breast cancer Roberto Salgado a,b,∗ , Philippe Aftimos b,c , Christos Sotiriou b,c , Christine Desmedt b a

Department of Pathology, GZA Antwerp, Belgium Breast Cancer Translational Research Laboratory, Institut Jules Bordet, Université Libre de Bruxelles, Brussels, Belgium c Department of Medicine, Medical Oncology Clinic, Institut Jules Bordet, Université Libre de Bruxelles, Brussels, Belgium b

a r t i c l e

i n f o

Keywords: Multifocal breast cancer Breast cancer subtyping Inter-lesion heterogeneity Clinical trials

a b s t r a c t The 7th edition of the TNM defines multifocal breast cancer as multiple simultaneous ipsilateral and synchronous breast cancer lesions, provided they are macroscopically distinct and measurable using current traditional pathological and clinical tools. According to the College of American Pathologists (CAP), the characterization of only the largest lesion is considered sufficient, unless the grade and/or histology are different between the lesions. Here, we review three potentially clinically relevant aspects of multifocal breast cancers: first, the importance of a different intrinsic breast cancer subtype of the various lesions; second, the emerging awareness of inter-lesion heterogeneity; and last but not least, the potential introduction of bias in clinical trials due to the unrecognized biological diversity of these cancers. Although the current strategy to assess the lesion with the largest diameter has clearly its advantages in terms of costs and feasibility, this recommendation may not be sustainable in time and might need to be adapted to be compliant with new evolving paradigms in breast cancer. © 2014 Elsevier Ltd. All rights reserved.

1. Introduction Multifocal breast cancer, defined as the presence of multiple tumoral lesions in the same breast, is a relatively poorly explored area compared to unifocal breast cancer. The reported incidence varies between 6% and 77% [1,2], with the largest study so far (n = 8935) reporting an incidence of 20.8% [3]. Generally, it has been associated with an increased risk of regional lymph node metastases as well as with an increased risk of local relapse and potentially a worse outcome [4,5]. The importance of intrinsic breast cancer subtyping and the emerging awareness of intratumoral heterogeneity in breast cancer have led to paradigm shifts in breast cancer biology. The applicability of these principles to multifocal breast cancer is unclear. Here, we will provide an historical overview of multifocal breast cancer definition and characterization. We will further challenge the current clinical approach based on recent insights related to the clinical and molecular characterization of this non-negligible subset of breast cancer patients.

Abbreviations: MC, multicentric; MF, multifocal. ∗ Corresponding author at: Breast Cancer Translational Research Laboratory, Boulevard de Waterloo 121, Institut Jules Bordet, 1000 Brussels, Belgium. Tel.: +32 2 541 35 34. E-mail address: [email protected] (R. Salgado). http://dx.doi.org/10.1016/j.semcancer.2014.07.002 1044-579X/© 2014 Elsevier Ltd. All rights reserved.

2. Current anatomic pathology concepts of multifocal breast cancer 2.1. Definition of multifocal breast cancer The large variability in reported incidence of multifocal breast cancer [1,2] may partly be explained by several factors: first, the different sensitivities of the imaging-method used in the preoperative setting; second, the degree of sampling by the pathologist at macroscopic examination, since it has been demonstrated that a more thorough sampling resulted in the identification of additional lesions [6]; third, the lack of awareness of the more than rare occurrence of multifocal breast cancer; and finally, the different definitions that have been used in the past to denominate multiple breast cancers arising in the same breast. The definition of multifocal breast cancer has evolved over the years. Traditionally, multiple synchronous breast cancer lesions have been defined as either multifocal or multicentric, depending on whether the lesions are located in the same or in different quadrants, respectively. The use of breast cancer quadrants for classification purposes is however now recognized as inappropriate since it is based on an arbitrary definition, which does not match the anatomy of the breast. Indeed, the breast develops in a continuous fashion forming a complex network of radially arranged breast ducts connecting the mammary lobules with the nipple. Also,

112

R. Salgado et al. / Seminars in Cancer Biology 31 (2015) 111–118

from the blood vessel distribution, it is clear that dividing the breast into 4 quadrants has no anatomical correlate [7]. The arterial blood supply to the breast arises from the internal thoracic artery, with associated blood supply from the intercostal and lateral thoracic arteries. The venous anatomy parallels the arterial anatomy. The lymphatic network in a normal breast, besides being developed in parallel with the arterial blood supply, has also extensive communicating channels between the superficial and deep lymphatic vessels. The branching and communicating lymphatic pattern, the corresponding superficial and deep lymphatic flow also illustrate that the drainage to the axillary lymph nodes does not necessary need to pass through the sub-areolar plexus but can be derived directly from deeper lymphatic structures, adding more evidence that the boundaries of a normal breast cannot be divided into 4 quadrants [8]. Considering the arbitrary distinction between multifocality and multicentricity, multiple simultaneous ipsilateral and synchronous lesions are now only defined as multifocal breast tumors, in the latest edition of the TNM (Tumor-Node-Metastasis) classification [9], provided they are macroscopically distinct and measurable using current traditional pathological and clinical tools. In this review we will thus further refer to these lesions as multifocal breast cancer, irrespective of the localization of the lesions. 2.2. Evaluation of standard clinico-pathological variables in multifocal breast cancer The TNM classification does not include biological variables such as for example the estrogen receptor (ER), progesterone receptor (PR), the HER2 oncogene or more recent biological variables, which help to identify which patient does need treatment (=prognostic information) and/or which treatment is best suited for the patient (=predictive information). It can be debated in the current era of increased use of targeted systemic treatment, whether additional biological characterization of the disease may become more important than mere anatomical information. The pathological diagnosis of breast cancer relies on the evaluation of the histological grade; the TNM stage together with the assessment of ER, PR, HER2 and potentially Ki67. All this information is subsequently integrated in scoring systems that allow accurate estimations of the probability of recurrence and death from breast cancer [10]. The decision to treat a patient depends on both the risk of relapse on the one hand and estimated predicted benefit to a particular treatment. Breast cancer has also been categorized into distinctive intrinsic subtypes [11]. These subtypes have been associated with distinct biological characteristics, metastatic evolution patterns, and consequently with different treatment strategies [10]. Moon et al. have investigated the prognostic impact of multifocal breast in association with the different molecular subtypes. There was a subtype-specific prognostic importance of multifocal breast cancer in overall survival, with triple-negative breast cancer patients having the worst survival. Importantly, they also show that the incidence of lymphatic metastasis is incremental in triple negative subtypes reaching a plateau in incidence of nodal metastasis only in tumors OS-MC (p = 0.001)

–

–

BCS-UF > BCS-MF (p < 0.001)

–

[21]

288 (5%)

DFS-UF > DFS-MF (p < 0.001) RFS-UF > RFS-MC-MF (MF p = 0.007; MC p = 0.019) –

[28]

147 (11%)

No

No

Higher in MF (local relapse p < 0.001, metastases p < 0.003) –

[29]

101 (18%)

–

–

–

[23]

906 (24%)

–

–

–

–

No difference in LR recurrence No difference in LR after BCT

Lower survival when phenotype heterogeneous versus homogeneous for MF

[3]

[33]

1398 (15.6%)

1187 (6%)

[18]

110 (100%)

–

[26] [34]

202 (15%) 342 (9%)

No –

[56]

[2]

153 (34%) (MF and diffuse)

942 (24%)

–

–

–

–

– –

– Higher in MF

Lower survival for MF (p = 0.0016)

–

–

Worse RFS for MC (p = 0.02)

Lower BCSS for MC (p = 0.01)

–

–

No

No

No

Higher in MF in the univariate analysis

Lower survival for MF or diffuse (p = .02) –

–

–

–

Higher in MF

–

–

60 (21%)

[57]

148 (30%)

–

[58]

225 (44%)

–

[30]

94 (11%)

–

[31]

94 (11%)

–

Lower survival for MF (p = 0.022) Lower breast cancer specific survival for MF –

[60] [27]

158 (17%) 141 (17%)

– –

No No

– –

[62]

149 (14%)

–

–

[61] [63]

945 (13%) 58 (19%)

Lower DFS No

No No

Higher when gross MC – –

1554 (6%)

–

–

[25]

[59]

Higher in MF (T1 p = 0.05, T2 p = 0.003) Higher in MF (p < 0.0001)

–

–

–

–

Higher in MF (p = 0.007) – – – – –

Comment

Same RFS and OS when patients guideline-adherent

No difference in nodal status when sum of diameters is considered for TNM No difference in nodal status when sum of diameters is considered for TNM BCT safe for MF/MC tumors

Women aged 50–69 years old with small (25. This trial does not exclude patients with synchronous bilateral breast cancer nor patients with multifocal or multicentric breast cancer. In light of the data presented earlier [40], sampling of all foci should have been mandatory since the recurrence score can vary from one lesion to the other. This also applies to the RxPONDER trial (NCT01272037). The aim of this trial is to determine if the recurrence score (Oncotype DX) could help discriminate patients with ER positive early breast cancer and 1–3 involved lymph nodes that could avoid adjuvant chemotherapy. The inclusion/exclusion criteria of this trial differentiate between multicentric (more than one invasive cancer ≥2 cm from the largest lesion within the same breast quadrant or more than one lesion in different quadrants) and multifocal (more than one invasive cancer