143 / 220
concurrent END. The end-points of this study were SN
identification rate, false-negative rate (FNR) and disease-
free survival (DFS) at 3 years post-recruitment. Because
of the lack of contemporaneous SNB data, we have used
comparable data from patients treated by conventional
END as well as SNB data in similar tumour groups who
routinely use SNB in management of the neck to con-
textualise the results. The aim of this investigation was to
assess whether SNB is a safe and reliable therapeutic
technique in T1
e
T2 oral squamous cell carcinoma.
2. Patients and methods
A European multicentre prospective study (October
2005
e
October 2010) was approved by the European
Organisation for Research and Treatment of Cancer
and local ethics committee, with patients providing
informed consent. Eligible patients had 0.5- to 4-cm
squamous cell carcinoma with an N0 neck on CT and/or
MRI (
<
1.1 cm or up to 1.5 cm in level II and no atypical
features)
ultrasound-guided fine needle aspiration
cytology. SENT was principally designed for oral
squamous cell carcinoma; however, tumour-bordering
structures of the oropharynx which were transorally
accessible and resectable (without mandibular split or
robotic techniques) were also included. Tumour location
was recorded according to Systematized Nomenclature
of Medicine (SNOMED) topography code
[14]
apart
from ‘oral tongue posterior 1/3’ which related to tu-
mours of the posterior part of the oral tongue and not
tongue base tumours.
Patients with a previous malignant neoplasm of the
head and neck or any disease that might have altered
lymphatic drainage were excluded. Patients had to be fit
enough to tolerate a completion neck dissection if the
SNB proved positive.
A total of 480 cases were recruited prospectively from
14 European centres. The criterion for unit participation
was completion of at least 10 successful training SNB
procedures (validated against neck dissection) prior to
recruiting to SENT. Sixty-five patients (14%) were
excluded from the final analysis (
Table 1
).
When adjuvant treatment (radiotherapy [RT] or
chemoradiotherapy) was given during the follow-up
period for close margins or a metachronous primary
tumour (N
Z
17), patients were excluded on the premise
that radiation fields extended into the upper neck and
could theoretically extinguish missed metastases,
thereby erroneously reducing the SNB FNR.
Pre-operative lymphoscintigraphy was performed
within 24 hours of surgery after Tc-99m nanocolloid
(Nanocoll/Nanocis ) was injected using a standardised
technique
[15]
at four points around the tumour (median
dose 57 MBq
e
interquartile range 60 MBq).
The position of the SNs was marked on the neck. At
surgery, the SNs were detected by a hand-held gamma
probe and in 164 of 415 patients (39%), peritumoural
injection of blue dye was given (SN recorded by colour,
radiation count and site in neck). Lymph nodes with
radiation count more than three times the background
activity were considered SNs. If a radiation hot spot was
in more than one neck level (SN versus second or third
echelon nodes), then the primary SN was decided by
maximum radiation count.
The SNs were fixed in 10% neutral-buffered formalin
and a validated protocol for analysis was followed
[16]
.
Five serial sections were cut every 150
m
m through the
block and one from the centre of each series was stained
with haematoxylin and eosin (H&E). If metastasis was
still not detected, an adjacent section at each level was
stained with anti-pan cytokeratin antibody AE1/3. If
cytokeratin was detected but the viability of the cells was
in question, the adjacent serial sections were examined
stained with H&E. One center (67 in 415: 16% of pa-
tients) cut a single frozen section (FS) from the midline
of the node, with remaining specimen examined as
above. This allowed on-table diagnosis
[17]
and imme-
diate neck dissection if the FS was positive.
SENT recorded metastasis as viable or non-viable
deposits sized in terms of percentage of the total node.
For the purposes of this report, SNB
þ
nodes were
retrieved where possible (75 in 94 cases, 80%) and re-
graded according to the Union for International Cancer
Control (UICC) Seventh Edition guidelines
[18]
.
Deposits were re-classified as isolated tumour cells (ITC,
<
200 cells or
<
0.2 mm deposit with no stromal reac-
tion), micrometastasis (0.2
e
2 mm), and macrometa-
stasis (
>
2 mm). In the SENT cohort, ITC was treated as
a positive neck (completion neck dissection performed
within 3 weeks).
Tumours were excised aiming for pathological clear
margin of
>
4 mm, and all defects were closed without
free flap reconstruction. Neck specimens were pinned
out maintaining alignment and fixed in neutral-buffered
formalin. They were examined macroscopically and by
routine H&E with cervical metastasis mapped by neck
Table 1
Cases excluded from SENT trial.
Reason for exclusion
Cases
% Total recruited
(n
Z
481)
Failed lymphoscintigraphy
N
Z
1 0.5
Failed identification of SN at
surgery
N
Z
5 1
Obvious nodal disease at surgery
N
Z
1 0.5
Breach of protocol (neck dissection
despite negative SNB (n
Z
5),
or no neck dissection after
positive SNB (n
Z
20)
N
Z
25 5
Lost to follow-up
N
Z
16 3.5
Radiotherapy for close margin or
second primary tumour
N
Z
17 4.5
Total
N
Z
66 14
SENT, Sentinel European Node Trial; SN, sentinel node; SNB,
sentinel node biopsy.
C. Schilling et al. / European Journal of Cancer 51 (2015) 2777
e
2784
123