Assessment of breast cancer response to neoadjuvant chemotherapy: Is volumetric MRI a reliable tool?

Abstract

The purpose of this study was to evaluate the reliability of volumetric magnetic resonance imaging (MRI) in breast cancer size assessment before, during and after neoadjuvant chemotherapy (NAC).

Volumetric MRI measures performed on 15 patients with breast cancer were compared with volumes reckoned upon mean lesional diameters, using the same MRI data. Concordance correlation coefficient (CCC), Bland & Altman plots, RECIST evaluation and Cohen's Kappa were assessed, to evaluate the agreement between the two methods.

CCC was computed before (0.9357), during (0.8053) and after (0.7499) NAC, in all examinations pooled together (0.8617), and on final tumor volume as a percentage of baseline volume (0.9224). In 2/15 (13.3%) cases RECIST assessment was different. Cohen's Kappa was 0.787 (CI_95% = 0.513–1.062).

In summary, volumetric MRI is a reliable tool to assess breast cancer size before, during and after NAC. Further investigations are needed to understand whether improvements in surgical planning are feasible.

Introduction

Nowadays it has been fully elucidated that neoadjuvant chemotherapy in breast cancer treatment is as effective as adjuvant therapy [1]. Moreover, neoadjuvant chemotherapy enables an increased rate of conservative surgery [2], that, associated with radiation therapy, guarantees overall survival and disease free survival rates similar to those obtained by mastectomy [3]. In addition, several patients inoperable at diagnosis, as a consequence of tumor's shrinkage can subsequently undergo radical intervention. Furthermore, unlike adjuvant therapy, neoadjuvant therapy allows direct assessment of cancer response to treatment, which correlates with patients’ survival [4].

In this contest, the role of imaging is of primary importance, in order to tailor the therapeutic plan upon patients’ needs and to achieve the minimal effective therapy. An accurate staging at diagnosis is important both for treatment monitoring and for determining patients’ prognosis [5]. Early identification of non-responders could result in sparing them ineffective and toxic treatments, and in providing them with a more effective regimen or with surgery. Finally, accurate assessment of residual disease after chemotherapy is needed in order to plan surgery [6].

Several studies showed that magnetic resonance imaging (MRI) is more accurate than any other imaging technique in preoperative assessment [7].

After neoadjuvant treatment, chemotherapy-induced changes in breast cancer – mainly necrosis and fibrosis – may interfere with clinical examination and with conventional breast imaging evaluation [8]. Because of its ability in distinguishing between fibrous and vascularized tissue, MRI is more accurate than X-ray mammography and sonography in assessing residual disease even though it has been reported that, compared with pathological findings, it usually overestimates large lesions, and underestimates the small ones [9], [10].

Whatever technique is used, breast cancer size is usually quantified by tumor's longest diameter measure, according to response evaluation criteria in solid tumors (RECIST) [11].

However, it has been recently suggested that when assessing lung lesion size changes – especially in case of irregular shape – a three-dimensional (3D) approach could be more accurate than a uni- or bi-dimensional one, possibly leading to an higher sensitivity in the detection of lung malignancies, that are expected to grow faster than benign lesions [12], [13]. The high sensitivity and usefulness of the 3D approach in identifying lesion size changes has been confirmed in breast imaging as well: Partridge et al. and Martincich et al. pointed out the volumetric-MRI capability of predicting both the degree of cancer response to neadjuvant therapy in a early stage of the treatment, and the rate of disease recurrence at the end of it [14], [15].

In this context, as lesion diameter is usually measured because it approaches lesion volume, and irregular modification of lesion shape – often occurring during neoadjuvant chemotherapy – could result, according to diameter measure, in an inaccurate evaluation of volume changes, we hypothesize that volume measures are at least as reliable as unidimensional or unidimensional-derived measures in assessing breast cancer size changes and, thus, in monitoring neoadjuvant therapy. Moreover, we hypothesize that tumor volume evaluation is at least as reliable as the traditional evaluation of tumor size, but it could improve surgical planning, in order to achieve the minimal effective therapy, because of the higher informative content given to the surgeon. Therefore, our purpose is to compare volume measures performed with a 3D-imaging software from MRI data, with volume measures reckoned from the mean lesional diameter from the same MRI data, before, during and after the neoadjuvant therapy.