Histopathology of the Sentinel Node
tency due to the personal opinion of the histopa-
thologist as well as difficulties in defining features
accurately. For instance, the grading system in the
breast is assessed by analysing three different fea-
tures [7]: the proportion of the tumour-forming
tubules (<10%, 10%– 75 %, > 75 %), the pleomor-
phism (difference in size and shape of the nuclei)
and mitotic count over ten high power fields (hpf).
The hpf will clearly be dependent on the field
diameter of the
¥
40 lens used to assess the slide.
To over co me t his problem, charts have been creat-
ed to standardise the cut-off values for the mitotic
count for a given field diameter. Although there is
an attempt to be objective in such a grading
system, it must be quite clear that the assessment
of the proportion of tumour showing good tubule
formation and the assessment of pleomorphism
are still subjective exercises. Variations in mitotic
count may also arise depending on which part of
the tumour is examined. Consequently, although
pathologists agree broadly on the type and grade
of tumours, the agreement is by no means perfect.
A number of studies have shown that
k
-statistics
for inter-observer variation for the assessment of
tumour type, tumour grade and tumour size are
lower than might be predicted without an under-
standing of the limitations of examination [8, 9].
Var iat ions in the histopathological assessment
may also result from the techniques used for the
examination of the specimen. Breast cancer is a
heterogeneous disease both clinically and mor-
phologically. It is not surprising, therefore, that
factors such as the number of sections examined
per case could influence the typing and the grad-
ing of the tumour. By examining a limited amount
of tissue, the proportion of tumour showing good
tubule formation,the degree of pleomorphism and
the mitotic count could all be underestimated.This
is occasionally seen when high-grade invasive
Background
Breast carcinoma is the commonest malignancy in
women and it is estimated that 1 in 12 women will
develop such a tumour within their lifetime. The
multistep model of breast carcinogenesis suggests
that tumours arise via clonal expansion with the
acquisition of multiple genetic hits that eventually
results in an ability to invade stroma and subse-
quently to produce metastatic disease [1]. As part
of this evolution, it has generally been assumed
that tumours first metastasise to regional lymph
nodes and subsequently to systemic sites. How-
ever, the alternative view, that breast cancer is a
systemic disease at the outset, has also been pro-
posed [2]; according to this model, breast cancer
cells are already present in the systemic circulation
before clinically evident lymph node metastases.
Although the concept of tumours first metastasis-
ing to lymph nodes may be an oversimplification,
lymph node status still remains the most impor-
tant prognostic indicator [3, 4]. Other important
prognostic indicators include tumour type, size
and grade [5, 6].
The aforementioned four prognostic features
are routinely assessed during the histopathological
examination of a breast specimen. There are two
principal reasons why the data derived from such
an examination may be inaccurate:
1. Inter-observer variability in the assessment of
the histopathological parameters
2. Inadequacy in the techniques used for the ex-
amination
The lay public, as well as many doctors, has dif-
ficulty in understanding how examination under
the microscope may not reveal an accurate diagno-
sis. The histopathological examination of tissues is
a subjective assessment and prone to inconsis-
chapter
9
