146 / 220
been missed by conventional treatment of ipsilateral
neck dissection.
In head and neck cancer, historically, there has been
concern that biopsy of suspected neck metastasis
would facilitate dissemination of tumour in the neck.
A systematic review
[20]
of 109 papers calculated
regional recurrence rates of 13% in surgically treated
early-stage oral cancer. A further review of 164
[21]
patients with pT1
e
T2 tongue SCC staged pN0 after
END reported a regional recurrence rate of 18%. The
results of SENT when reported in an identical
way show the neck recurrence rate for SNB and
SNB
þ
and the total group were 5%, 15% and 7.5%.
The low rate of regional recurrence argues against
SNB causing tumour spillage and in turn neck
recurrence.
Two- and 5-year overall survival in early oral and
oropharyngeal carcinoma is in the region of 82% and
76%, respectively
[20,22]
. In this study, overall crude
(88%) and DSS (94%) are unlikely to change signifi-
cantly and suggest strongly that SNB does not
adversely affect outcome. An FNR of 14% is similar to
that reported in a meta-analysis of 25,000 melanoma
patients (12.5%)
[23]
and 20% FNR in 10-year follow-
up of the Multicenter Selective Lymphadenectomy
Trial (MSLT) trial in melanoma
[24]
. However, this is
on the borderline of acceptability and we should aim
to reduce this to the 7% FNR accepted in breast
cancer
[25]
. Further analysis of the factors associated
with a false-negative biopsy is warranted but initial
review of our data suggests that operator factors are
principally responsible for the FNR. It is well estab-
lished that there is a learning curve to the SN tech-
nique
[11]
.
It is of particular note that previous studies
[9,11]
indicated that SNB was less reliable for tumours in the
floor of mouth presumably due to the close proximity of
the injection site to the primary draining nodes. The
same association was not found in this study.
The major positive patient benefit of SNB is that in
this series 71% of patients were spared neck dissection
with consequent improved function and reduced
morbidity
[26,27]
. There were also 47 patients with
midline tumours who by convention would have
received bilateral neck dissection. In this group, only
eight underwent bilateral and eight unilateral dissection
based on positive SN. A low complication rate as well as
Fig. 1. Overall survival for SNB
þ
versus SNB biopsy
(p
Z
0.00083). SNB, sentinel node biopsy.
Fig. 2. Overall survival for 0, 1, or 2 positive sentinel nodes
(p
Z
0.000016).
Fig. 3. Overall survival by metastasis type: isolated tumour cells
(I) versus macrometastasis (Ma) versus micrometastasis (Mi)
(p
Z
0.0318).
C. Schilling et al. / European Journal of Cancer 51 (2015) 2777
e
2784
126