] Although the basic principles
did not change, stages involving clinical and pathological
prognoses were added in addition to the anatomic
stage. There were no changes in the anatomic stage for
breast cancer in the first and the updated manuals.
This review aims to reveal the changes in the stages
of our institutional breast cancer patients according to
the first and the updated versions of the AJCC manuals
and to compare the clinical reflections with the help of
stage studies involving the updated manual.
Stage Changes According to the First and Updated
Versions of the AJCC Manual (8th edition)
Here, 353 patients with stages I-III breast cancer who
underwent surgery as the primary treatment at the Istanbul
University, Institute of Oncology between 2004 and 2006 were considered. Data involving T, N, M, ER,
PR, HER2, and tumor grade were noted in accordance
with the section on breast cancer of the AJCC manual
(8th edition). The histological grade of the tumor was
evaluated according to the Elston-Ellis modification
of the Scarff-Bloom-Richardson grading system.[]
The ER biomarkers were determined by immunohistochemistry
and were recorded as the percentage of cells
stained as positive. The cut-off of 1% was used for the
treated patients, which was consistent with the changes
in the American Society of Clinical Oncology (ASCO)
guidelines.[] The PR status was determined by immunohistochemistry,
and staining of 1% of the cells
or more was considered to be positive for PR. HER2
status was defined as positive if the score was 3+ on
immunohistochemistry or fluorescence in situ hybridization,
demonstrating gene amplification.[] Because
the anatomic stage did not change in the first and the
updated versions of the AJCC manuals, all the patients
were staged according to the anatomic stage, prognostic stage as per the first version, and pathological prognostic
stage as per the updated version. Stage changes
(unchanged, upstaged, and downstaged) were detected.
The 21-gene Oncotype DX (Genomic Health, Inc.)
breast recurrence score contains both prognostic and
pathological prognostic staging systems. Oncotype DX
assay was not performed in this series. However, the
potential effect of the Oncotype DX multigene assay
was examined according to the first and the updated
versions.
The median age was 48 years (24-79 years). Tumor
grade was grade 1 in 19 patients (5.4%), grade 2 in 154
patients (43.6%), and grade 3 in 180 patients (51%). In
addition, 253 (71.7%) patients were detected as having
ER (+); 265 patients (75.1%), PR (+); and 64 patients
(18.1%), HER2 (+). The anatomic, prognostic, and pathological
prognostic stage results are listed in Table 1.
Table 1: Patient distribution at three different stages according to the first and the updated versions of the AJCC manual
(8th edition) for breast cancer staging
One of the most remarkable points is that in the
first version of the manual, 50 (14.2%) patients had no
prognostic stage; however, in the updated version, all the patients had a pathological prognostic stage. The
other point is that in the first and the updated versions,
the patients were downstaged as compared to
the anatomic staging system. However, downstaging
was more prominent in the updated version than that
in the first version. The degrees of stage migration are
listed in Table 2. Oncotype DX multigene analysis was
appropriate in 97 patients (27.5%). Oncotype DX assay
was not performed in this series. Regardless of the
Oncotype DX recurrence score, out of the 97 patients,
46 (47.4%) would still be in stage IA in the first version
and 83 (85.6%) would still be in stage IA in the updated
version (Table 3).
Table 2: Stage changes according to the first and the updated versions of the AJCC manual (8th edition) for breast cancer
Table 3: Oncotype DX recurrence score evaluation according to the first and the updated versions of the AJCC manual (8th
edition) for breast cancer
Focusing on Stage Changes with the Help of Latest
Studies
The prognostic stage in breast cancer was developed
using the National Cancer Database (NCDB) consisting
of approximately 238.000 patients" data, who were
diagnosed and treated between 2010 and 2011 and
whose clinical information can be accessed, including
AJCC TNM staging, tumor grade, ER, PR, and HER2
biomarkers. The prognostic subgroup and survival calculations
were performed according to the TNM stage,
tumor grade, ER, PR, and HER2 statuses. On the basis
of these analyses, 170 different prognostic groups-defined
on the basis of tumor biology, varying in TNM
stages-were distributed between the stages of 0 and IV.
At the end of 2016, the 8th edition of the manual was
published.[,]
After the publication of this update, the prognostic
stage was validated in additional cohorts.[-] The anatomic and prognostic stage analyses of the updated
manual in breast cancer subgroups was performed
on five of these studies [-,] and the general
group involving non-metastatic invasive breast cancer
patients was analyzed on three of them.[,,]
Hu et al. investigated the stage changes according
to the anatomic and prognostic stage systems in the
AJCC manual (8th edition).[] Weiss et al. reported
that patients with unknown prognostic stages were
not included in the survival analysis, although no
number of patients were specified in the literature.[]
Lee et al. reported that there were 497 patients (6.7%)
who did not have prognostic stage outputs while having
anatomical stage outputs.[] Hu et al. reported
that there were 77 patients (9.8%) who did not have
prognostic stage outputs.[] The prognostic stages of
50 (14.2%) patients could not be determined in our
institutional series.
In December 2017, the section on breast cancer
of the AJCC manual (8th edition) was updated in the
light of new data. Two analyses were performed on
the NCDB data. In the first case, the clinical information
of all the patients who showed the clinical prognostic
stage was used. In that study, 334.000 patients
who were diagnosed with invasive breast cancer between
2010 and 2012 and the median follow-up period
of 41.7 months were evaluated. All the patients
were included in this analysis, regardless of the treatment,
because most of them received systemic treatment
on the basis of their stages and biomarkers. The
clinical prognostic stage was recommended for all the
patients.
The second analysis was limited to the patients with
surgical resections as the initial treatment. Patients who
received preoperative systemic therapy or radiotherapy
(neoadjuvant therapy) were excluded from this analysis;
all the remaining patients were included, regardless
of the treatment received by them after surgery. Approximately
306.000 patients whose median follow-up
period was 42.3 months and diagnosed between 2010
and 2012 were included in this analysis. Pathological
prognostic staging should be calculated in patients undergoing
surgery as the initial treatment. Pathological
prognostic stage groups were established by combining
the anatomic stage groups with the grade, ER, PR, and
HER2 statuses. Here, 120 different patient categories
were revealed in this way.
In the first version as well as in the updated version
of the AJCC manual (8th edition), patients in T1-
2, N0, M0, grades 1-3 and those with ER (+), HER
2(-), and Oncotype DX recurrence score <11 were
classified as stage IA.[] In almost all the studies, Oncotype
DX multigene assay results were reported as
"does not exist." In our institutional series, there were
97 (27.5%) patients appropriate for Oncotype DX determination.
If the Oncotype DX recurrence score of
all the patients was <11, all of them would be identified
as stage IA. If it was ≤11, 46 (47.4%) out of the 97
patients would be in stage IA, according to the first
version; similarly, the remaining 30 (30.9%) patients
would be in stage IB; 16 (16.5%) patients, stage IIA; 1
patient, stage IIB; and 1 patient, stage IIIA. Further,
3 patients with no prognostic stage would not have
been able to undergo staging. If this value was ≤11,
83 (85.6%) out of the 97 patients would be in stage IA
according to the first version; similarly, the remaining
12 (12.4%) patients would be in stage IB, and 2
(2.1%) patients, stage IIA. In the pathological prognostic
staging group, the number of stage IA patients
increased to 38.2%. Although it is necessary to validate
these results with further studies in the future,
the number of patients who should be examined with
Oncotype DX test according to pathological prognostic
staging seems to have decreased.
By analyzing the anatomic and prognostic stages
according to the first version of the AJCC manual
(8th edition), Hu et al. reported that the patient distribution
rate was as follows: 24% patients in stage IA;
31%, stage IIA; 20%, stage IIB; 13%, stage IIIA; 0.9%,
stage IIIB; and 8.5%, stage IIIC.[] In our study, the
anatomic stage distribution was similar: 27.8% of
them were in stage IA; 28%, stage IIA; 20.1%, stage
IIB; 15%, stage IIIA; 0.8%, stage IIIB; and 7.6%, stage IIIC. The prognostic stage distribution was examined
according to the first version. Hu et al. reported that
the distribution was 15.9% in stage IA; 27%, stage IB;
6.9%, stage IIA; 11.6%, stage IIB; 10.5%, stage IIIA;
8.5%, stage IIIB; 7.1%, stage IIIC; and 9.8%, unknown
stage.[] While in our study, these values are 14.4%
in stage IA; 19.5%, stage IB; 13.6%, stage IIA; 7.1%,
stage IIB; 7.1%, stage IIIA; 13%, stage IIIB; and 14.2%,
unknown stage. Considering pathological prognostic
staging (updated version of the AJCC manual), there
was an increase of 8.7% and 22.1% in stage IA, 21.2%
and 2.3% in stage IB, 13% decrease and 1.4% increase
in stage IIA, 9.3% decrease and 3.7% increase in stage
IIB, 6.2% decrease and 1.7% increase in stage IIIA,
3% increase and 9.2% decrease in stage IIIB, and 5.3%
decrease and 8.7% decrease in stage IIIC according
to the anatomical staging and first version of AJCC
manual (8th edition), respectively. In particular, according
to the anatomic stage, an increase in stage IB
was detected, whereas an increase in stage IA was also
detected according to the first version of the AJCC
manual (8th edition).
In our institutional series, it is revealed that downstaging
increases in relation to the pathological prognostic
staging system. In patients who require Oncotype
DX multigene analysis, it can be emphasized that
it would still be stage IA at the rate of 85.6%, independent
of the Oncotype DX multigene score.
