The American Joint Commission Cancer 8th Edition Prognostic Stage Including Oncotype DX® Recurrence Score: Impact on Staging of Early Breast Cancer

The recently published American Joint Commission on Cancer (AJCC) Cancer Staging Manual, 8th edition [1], introduced into United States clinical practice in January 2018, proposes a new breast cancer (BC) Prognostic Stage (PS) that combines histological tumour grade, oestrogen receptor (ER), progesterone receptor (PgR), and human epidermal growth factor-2 (HER2) receptor status with traditional Anatomic TNM Stage (AS) [2, 3]. In a further modification, patients with ER-positive, HER2-negative, lymph node (LN)-negative BC, and an Oncotype DX® recurrence score (RS) of < 11 are assigned to PS 1A irrespective of histological grade and tumour size up to 5 cm. The changes have been proposed in recognition of the prognostic influence and therapeutic effects associated with these parameters and in attempt to improve personalised patient care. These innovations represent a major departure from traditional practice and a move away from the current harmonised BC staging system agreed by AJCC and Union of International Cancer Control (UICC) [4, 5].

This study assesses the impact on BC staging using the newly proposed AJCC PS, including that modified by low Oncotype DX® RS, compared with the traditional AJCC/UICC AS, in a population of patients with early primary invasive BC and RS < 11.

The study population comprised patients with a primary invasive BC Oncotype DX® RS < 11. The population was identified from a large consecutive series of patients with ER-positive, HER2-negative, LN-negative, primary BC whose tumours were tested using the Oncotype DX® multigene assay between November 2011 and December 2016. Patients presented through the Irish National Breast Screening Programme (BreastCheck) or through the symptomatic service and were diagnosed and treated at St. Vincent’s University Hospital, Dublin, Ireland.

Therapeutic BC excision (wide local excision or mastectomy) and sentinel and non-sentinel LN specimens were prepared, dissected, and blocks selected according to Royal College of Pathologists (RCPath), United Kingdom (UK) macroscopy protocols [6]. Tumours were reported in accordance with RCPath BC minimum datasets [6] that take account of determination of invasive tumour size as defined by AJCC/UICC 7th editions [4, 5, 7] and histological tumour grade using the Nottingham method [8]. All LNs were negative in the study group in accordance with study inclusion criteria.

ER status was determined using immunohistochemistry in accordance with standardised protocols [6]. HER2 status was determined using first-line IHC and reflexive dual-probe HER2/CEP17 fluorescent in situ hybridisation (FISH) methodology in equivocal cases as previously described by our group [9]. During the period of study, PgR studies were not routinely performed on tumours that were strongly ER positive. Information on PgR status was derived from the Oncotype DX® report.

To qualify for Oncotype DX® testing, the pathological size of the primary tumour must be categorised as T1c (1.1–2.0 cm), T2 (2.1–5 cm), or T1a (< 0.5 cm) or T1b (0.5–1.0 cm) with adverse histopathological features, e.g. histological grade 2 or 3 or the presence of lymphovascular invasion.

AS was assigned according to the AJCC Classification of Breast Cancer Staging, 8th edition (unchanged from the 7th edition). PS and PS modified by RS < 11 were assigned according to the AJCC 8th edition.

Traditional AS was compared with PS and with Oncotype DX® RS-modified PS. Oncotype DX® RS-modified PS was also compared with PS alone.

All patients had ER-positive, HER2-negative, and LN-negative primary invasive BC in line with the selection criteria for Oncotype DX® testing. 127 of 729 patients had RS < 11 and comprised the study group. The mean patient age was 58 years (35–78). The mean tumour size category was 13.6 mm (5–50 mm). 126 patients had PgR-positive tumours. Pathological tumour size was T1a (n = 2), T1b (n = 19), T1c (n = 79), and T2 (n = 27). Histological tumour grade was grade 1 (n = 14), 2 (n = 98), and 3 (n = 15). The clinicopathological characteristics are summarised in Table 1.

Applying traditional AS, 100 patients were assigned to BC stage IA (T1N0), and 27 patients to stage IIA (T2N0). Applying PS (without Oncotype DX® RS), 89 patients were assigned to stage IA, 33 to stage IB, 4 to stage IIA, and 1 to stage IIB. Applying PS modified by Oncotype DX® RS of < 11, all patients were assigned to stage IA.

Comparing PS alone to traditional AS, 26.7% of patients (n = 34) underwent a change in stage, 9.4% (n = 12) to a higher stage, and 17.3% (n = 22) to a lower stage. Comparing PS modified by RS < 11 to traditional AS, 21.3% (n = 27) of patients were downstaged. Comparing modified PS to PS alone, 29.9% (n = 38) of patients were downstaged. The relative distribution of stage is shown in Table 2.

BC has traditionally been staged based on anatomic extent using the TNM classification system specified and agreed by the AJCC and UICC [4, 5]. Anatomic TNM provides information on prognosis, assists decision making regarding adjuvant therapy, provides selection criteria for patient entry to clinical trials and profiles BC epidemiology on a global level. It is relatively simple to apply and is easily understood. However, recent advances in our understanding of the impact of biological indices on prognosis and likely response to therapy have highlighted shortcomings in a purely anatomic-based staging classification system. In an attempt to address this deficit, the AJCC 8th edition has developed a PS that combines histological tumour grade, biomarker status, and the results of the Oncotype DX® 21 multigene assay, in a subset of patients, with AS to more accurately stratify BC patients according to prognosis and to further facilitate personalised treatment planning [1-3]. The PS was developed based on a review of the current literature and on an unpublished analysis, by one of the AJCC breast panel members, of 238,265 women diagnosed in 2010 who were included in the US database. Application of PS to that patient series resulted in an overall re-assignment of stage, to a worse or better category, in 41% of patients with BC. Recently published studies comparing AS with PS in varied patient subgroups have reported up to 64.4% stage change [10, 11].

One of the major criticisms of the AJCC 8th edition-proposed modifications to BC stage is the disparity between AS and PS and the potential for loss of information internationally due to variable access to biomarker studies [12]. Comparing AS to PS in the current study of 127 patients with the same ER, HER2, and LN profile, 34 patients (26.7%) underwent a change in stage. Twelve patients (9.4%) were upstaged and 22 (21.3%) were downstaged. In the group of patients who were upstaged, 11 of 12 moved from AS IA to PS IB due to histological tumour grade 3. One patient was upstaged from AS IIA to PS IIB due to negative PgR status. The impact of histological grade on PS underscores the prognostic power of Nottingham tumour grade independent of biomarker and LN status, previously validated in several studies [13, 14]. Tumour grade is a key component of the Nottingham Prognostic Index that has both prognostic and predictive value in BC management [15-17]. We have recently reported a strong correlation between the Nottingham Prognostic Index and Oncotype DX® testing in identifying low-risk patients [18]. In the current study, no patient was downstaged solely due to histological grade. This reflects the relatively low-risk tumour profile of the study group that specified ER-positive and HER2-negative biomarker profile and negative LN status.

All 22 patients who were downstaged moved from AS IIA to PS IB. This was due to the effects of positive hormone receptor (ER and PgR) and negative HER2 biomarker profile in conjunction with histological grade 1 or 2 and negative LN status. Positive hormone receptor status strongly correlates with response to adjuvant endocrine therapy and is also an independent favourable prognostic indicator, particularly in the first 5–7 years following diagnosis [19]. Positive PgR receptor status is associated with improved overall and disease-free 5-year survival, with a more pronounced effect in low-grade tumours [20]. Positive PgR status is also highly predictive of a low Oncotype DX® RS [21, 22]. HER2 positivity is usually observed in high-grade invasive BC and, in the absence of treatment, is associated with an adverse outcome [23, 24]. Inclusion of HER2 status in PS reflects the adverse impact of positive status on otherwise low-risk tumours but the overall improved prognosis of patients with HER2-positive tumours due to the development of trastuzumab and other anti-HER2-based chemotherapeutic regimes [25]. The impact of biological indices (grade and biomarker profile) and a low Oncotype DX® RS is further highlighted by consideration of the fact that a patient with tumour pathological stage pT2N0 equal to AS IIA may be assigned to any one of five PS categories (IA, IB, IIA, IIB, IIIA) depending on the results of these parameters.

The AJCC acknowledges the likely impact on staging of multigene tests and does not endorse any particular assay. The decision to incorporate the results of the Oncotype DX® 21 multigene assay [26-30] into PS in a selected subset of patients is supported by level 1 evidence, available at the time of writing, from the TAILORx trial (Trial Assigning IndividuaLised Options for Treatment [Rx]). The trial, sponsored by the National Cancer Institute (NCI) and coordinated by the Eastern Cooperative Oncology Group (ECOG), reported 99.3% distance relapse-free survival, 93.8% invasive disease-free survival, and 98% overall survival at 5 years in patients with a low Oncotype DX® RS (< 11) [30]. In the current study, 21% of patients (27 of 127), classified as AS IIA due to invasive tumour size T2 (> 20 and ≤50 mm), were reassigned to PS 1A due to RS < 11. All patients in the current study had ER-positive, HER2-negative, LN-negative tumours, not exceeding 50 mm, in line with TailorX selection criteria and our institutional requirements for Oncotype DX® testing. Reassignment to PS IA due to Oncotype DX® RS < 11 is restricted to this subset and is independent of PgR status.

While Anatomic TNM staging provides a more straightforward and easily applied system for classification of BC, the newly developed PS is arguably a more comprehensive approach. PS takes account of major recent advances in our understanding of BC at cellular level and more accurately predicts patient outcome [31-35]. A recent study of deceased patients found that PS separated patients according to overall survival time [36]. The value of incorporating biomarkers with pathological stage into a staging index to improve prognostication has previously been demonstrated by Yi et al. [37] at MD Anderson. This index was developed prior to the routine use of trastuzumab-based chemotherapy for patients with HER2-positive BC and highlights the need for regular updating of a so-called “living index.”

Difficulties with the proposed PS include the complex nature of the staging system with potential for lack of reproducibility [38]. Realisation that certain tumour profiles could not be assigned a PS has been addressed in a recent update [39]. More serious drawbacks include restriction of PS to patients who undergo surgery as their primary treatment in an era where many patients with HER2-positive and triple-negative breast tumours receive neoadjuvant chemotherapy. Disparity in staging and loss of information internationally will inevitably result from unequal access to biomarker studies. Within the more developed countries, variable access to multigene panels and economic considerations [40] will lead to differences in stage even with adoption of this new approach to BC. The availability of multiple genomic assays with varying concordance for individual patients is another confounding factor [41-45]. Inclusion of multigene assay results will necessitate regular updating and reformatting of modified PS as evidence of their clinical validity emerges [44, 46]. Improved understanding of BC heterogeneity at morphological and molecular level with implications for the efficacy of targeted therapy [47] will also need to be considered.

In conclusion, our study demonstrates that application of PS, including that modified by a low Oncotype DX® RS, results in tumour downstaging in approximately 20% of patients with early BC. In this series of patient with early BC, upstaging was seen in 9% of patients when staged according to PS and was primarily due to the impact of high histological grade. This new approach to BC staging aims to synthesise the anatomic extent and biological properties of a tumour such that the prognostic implication of a particular stage takes account of the likely response to personalised therapy [48]. Disadvantages include the potential for loss of epidemiological information internationally due to unequal access to biomarker and multigene assays and the need for regular updating as new predictive tests and evidence for their clinical validity emerges.

The authors have consulted the journal policy regarding compliance with ethical standards and state that accepted principles of ethical and professional conduct have been applied to this study.

The authors declare that they have no conflict of interest.