from the Ludwig Breast Cancer Study Group [29],
who showed detection of micrometastases in 9%
of their patients.Follow-up to 5 years showed a sig-
nificant difference in the disease-free survival and
the overall survival of these patients. The authors
concluded that the pathological examination of a
single H & E section is “probably no longer clini-
cally tenable”.Since then a large number of studies
have demonstrated that the detection of occult
metastasis within the lymph node has prognostic
significance for the patient,and it is quite clear that
there is a survival disadvantage for such patients.
So what are the implications for the routine ex-
amination of lymph nodes from breast cancer
specimens?
As mentioned above, it was been known for
more than 30 years that the examination of a single
section of a lymph node is inadequate for the
assessment of lymph node metastasis. Since the
lymph node status is an important prognostic
indicator, should we now abandon this technique
in favour of more extensive lymph node analysis?
The immediate problem that arises is how exten-
sive should this investigation be? In the studies
reported to date, there has been wide variability in
the level of sectioning of the lymph node, ranging
from a few levels through the node to complete
examination of the node using serial sectioning.
The majority of the pathology laboratories within
the United Kingdom and around the world would
come to a standstill if every single breast cancer
case were examined by cutting hundreds of serial
sections of every single lymph node that was
removed from the patient. Since extensive serial
sectioning is clearly not possible for most labora-
tories, what constitutes a reasonable number of
sections for the detection of most occult metas-
tases? Despite the vast number of studies reported
in the literature over the past two decades, the
answer to this question remains unclear, and the
majority of laboratories continue to use a single
H & E section to examine lymph nodes.
Use of Immunohistochemistry for the Detection
of Micrometastases
There have been at least 17 studies in the past
20 years reporting the use of immunohistochem-
istry in the diagnosis of occult metastases within
ductal carcinoma is present side by side with a
lower grade invasive lobular carcinoma.The prob-
lem has been further highlighted by the use
of Tru-cut biopsies for diagnosis. Occasionally,
examination of the resected specimen reveals a
marked difference in grade compared with that
predicted from the initial small biopsy. These limi-
tations, which are limitations of the techniques
used in the examination of tissues, are the princi-
pal sources of error in the assessment of lymph
node metastasis. Although the lymph node status
is the most important prognostic indicator, in
most laboratories the assessment of lymph nodes
is confined to examination of a single haematoxy-
lin and eosin (H & E) stained section from each
node. Is this type of lymph node examination
adequate and what are the consequences for the
patient?
Lymph Node Examination: Paraffin Sections
It was as long ago as 1948 that Saphir and Amromin
[10] demonstrated that serial sectioning of lymph
nodes improves the detection of lymph node
metastases. In their study, 33% of patients who
were initially diagnosed on routine histology as
being lymph node-negative were converted to
lymph node-positive by the examination of mul-
tiple serial sections. This was subsequently con-
firmed in 1961 by Pickren [11], who demonstrated
occult metastases in 22 % of node-negative cases.
Unfortunately no prognostic significance was
demonstrated between the two groups. Between
1971 and 1996, a further 29 studies of micrometas-
tasis have been reported in literature [12–40]. The
number of cases has ranged from 5 to 921 (Ludwig
Breast Cancer Study Group) and the type of exami-
nation has ranged from one haematoxylin and
eosin (H & E) section to serial sections of the entire
lymph node, with the use of immunohistochem-
istry for cytokeratins in some cases. The detection
rate for occult metastasis has ranged from 7% to
31%. The follow-up period for many studies has
been only 2–3 years; however, in one study, it was
greater than 16 years [38].It was not until 1987 that
Trojani et al. [23] demonstrated a statistically signi-
ficant difference in disease-free survival and over-
all survival in patients with such occult metastases.
The next study to confirm such an effect was that
chapter
9
104
