Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Oral Oncology
journal homepage: www.elsevier.com/locate/oraloncology
Department of Oral and Maxillofacial Surgery, University Medical Centre Utrecht, Heidelberglaan 100, G.05.129, PO Box 85500, NL 3508 GA Utrecht, The Netherlands
Department of Pathology, University Medical Centre Utrecht, Heidelberglaan 100, PO Box 85500, NL 3508 GA Utrecht, The Netherlands
a r t i c l e
i n f o
Article history:
Received 3 December 2013
Received in revised form 16 February 2014
Accepted 20 February 2014
Available online 14 March 2014
Keywords:
Oral squamous cell carcinoma
Local recurrence
Re-resection
Post-operative radiotherapy
Pathological margin status
Head neck cancer
s u m m a r y
Objectives: The treatment strategy of early stage oral squamous cell carcinomas (OSCC) resected with
close or involved margins is a returning point of discussion. In this study we reviewed the consequences
of re-resection (RR), postoperative radiotherapy (PORT) or watchful waiting (WW).
Patients and methods: Two-hundred patients with a primary resected Stage 12 OSCC of the tongue, oor
of the mouth and cheek were included and retrospectively analysed. Local recurrence ratio was related to
margin status, unfavourable histological parameters (spidery inltrative, peri-neural and vascular-invasive growth) and postoperative treatment modality. 3-year overall survival (OS) and disease-specic survival (DSS) was calculated in relation to margin status.
Results: Twenty-two of 200 (11%) patients had pathological positive margins (PM), 126 (63%) close margins (CM), and 52 (26%) free margins (FM). OS and DSS were not signicantly different between these
groups. Nine of 200 (4.5%) patients developed local recurrent disease. Two (9.1%) had a PM, ve (4.0%)
a CM and two (3.8%) a FM. Of the nine recurrences, ve patients had undergone PORT, one a RR, and three
follow-up. Watchful waiting for CM P3 mm with 62 unfavourable histological parameters showed,
besides margin status no signicant differences with the FM group.
Conclusion: With this treatment strategy, the local recurrence rate was 4.5%. No evidence was found for
local adjuvant treatment in case of close margins P3 mm with 62 unfavourable histological parameters.
Current data do not support the use of one treatment modality above any other.
2014 Elsevier Ltd. All rights reserved.
Introduction
For Stage 12 oral squamous cell carcinoma (OSCC) the preferred choice of treatment is complete surgical removal of the tumour. To achieve the best results in loco-regional control and
long-term disease-free survival, several authors believe that free
resection margins of at least 5 mm are essential [14], while others
disagree [57]. Complete removal should ideally be achieved at the
rst surgical procedure [2]. However, in 513% of resected early
OSCCs, microscopic tumour is present in the resection margin,
known as a positive margin [2]. A positive margin carries a high
612
Table 1
Denition of patient groups based on pathological margin status.
Group
Denition
n (%)
PM
22 (11)
CM
FM
126 (63)
52 (26)
modalities (i.e. RR, PORT and WW) were determined. For the
groups PM, CM and FM survival curves were calculated. Three-year
overall survival (OS), was calculated from the date of rst histological conrmation of OSCC to the date of death from any cause. The
three-year disease-specic survival (DSS) was the secondary outcome. For DSS-rates censoring occurred at the date of death from
causes other than OSCC or at the end of the follow-up period,
whichever came rst.
Patients and tumour characteristics in relation to the margin
status and the type of adjuvant treatment were analysed and compared. Patients in the WW group (with CM) were compared to patients in the FM group. Three-year OS- and DSS-rates were
determined for these two groups as well. Characteristics of the patients are reported as frequency (percentage) for categorical variables. Continuous variables are presented as mean (SD) when
normally distributed or median (range) when not. Gaussian distribution was conrmed by visual analysis of the histograms, QQ
plots and the ShapiroWilk test. One-way ANOVA was used for
hypotheses testing of normally distributed continuous data and
MannWhitney U test and KruskalWallis test for continuous data
that were not normally distributed. For categorical data P-values
were calculated with the use of Fishers exact tests. Using life table
techniques, DSS-rates and OS-rates were calculated, illustrated by
KaplanMeier plots. Covariates were compared with the log-rank
test. All test statistics were two tailed, and the signicance level
was set at p < 0.05. Survival analyses were performed with Stata
Statistical Software (Release 12. College Station, TX: StataCorp
LP) All other analyses were performed with the use of Statistical
Package for the Social Sciences (SPSS for windows, release 20.0.0
2011, SPSS Inc.).
Results
In the total cohort of 200 patients, 22(11%) had a PM, 126(63%)
a CM, and 52(26%) a FM. The three patient groups did not differ
regarding tumour site, patients habits, gender or age. Also, there
were no differences in the unfavourable histological parameters.
A signicant difference was found with relation to tumour diameter and thickness (p = 0.005 and p = 0.002 respectively): in the CM
group, tumours were bigger than those in the PM and FM group.
Table 2 shows all relevant data of the 200 patients included in
the study. Nine out of 200 (4.5%) patients developed recurrent disease at the primary site. Distribution over the different sub-sites is
shown in Table 4. Recurrences were located at the tongue in 6/105
(5.7%), the oor of mouth in 2/75(2.7%) and inside the cheek in 1/
22 (4.5%). In the PM, CM and FM groups, respectively two (9.1%),
ve (4.0%) and two (3.8%) patients had local recurrent disease. Of
the two patients with local recurrence in the PM group, one had
undergone a RR and one PORT. Of the ve local recurrences in
the CM group, four had received PORT and one was selected for
WW. Of all nine recurrences, ve patients had undergone PORT,
one a RR, and three developed during WW (Tables 3 and 4, Fig. 1).
Comparison of patients in the WW group (n = 77) with those of
the FM group (n = 52) showed, as expected due to selection, a signicant difference in margin status (p < 0.001) with a median
resection margin of 3.0 mm (range 15) in the CM group and
6.0 mm (range 5.110) in the FM group. However, no signicant
difference in development of local recurrence was encountered,
being 1/77(1.3%) in the WW and 2/52 (3.8%) in de FM group. Also
unfavourable histological parameters, thickness and diameter
were not signicantly different between these groups (Table 5).
The three-year OS was 91% (95% CI 6898%) in group PM, 87%
(95% CI 8092%) in group CM and 87% (95% CI 7493%) in group
FM (p = 0.86). The three-year DSS was 95% (95% CI 7299%) in
group PM, 91% (95% CI 8495%) in group CM, and 90% (95% CI
613
b
c
PM
n = 22
CM
n = 126
FM
n = 52
13 (59)
9 (41)
72 (57)
54 (43)
28(54)
24 (46)
P-value
Variables (n = 200)
PM
n = 22
CM
n = 126
FM
n = 52
2 (9.1)
5 (4.0)
2 (3.8)
0
1 (4.5)
1 (4.5)
4 (3.2)
1 (0.8)
0
2 (3.8)
0
0
0.91
b
Age yr
Mean
SD
Range
58.9
12.8
3183
63.3
12.6
2390
61.1
11.4
3686
9 (41)
13 (59)
67 (53)
59 (47)
26 (50)
26 (50)
0.21
0.57
a
12 (55)
10 (45)
73 (58)
53 (42)
30 (58)
22 (42)
11 (50)
9 (41)
2 (9)
67 (53)
46 (37)
13 (10)
27 (52)
18 (35)
7 (13)
Table 4
Distribution of recurrence in relation to margin status.
Table 5
Recurrence and pathological characteristics of group WW (with CM) and FM.
0.95
Variable (n = 129)
Local recurrence
Yes
No
Spidery growth no. (%)
Yes
No
Peri neural growth no. (%)
Yes
No
Angio invasive growth no. (%)
Yes
No
Tumour diameter (mm)
Median
Range
Tumour thickness (mm)
Median
Range
0.96
11.0
236
17.5
140
12.0
140
0.005
c
3.0
116
5.8
130
4.0
120
16(73)
7 (32)
3 (14)
81 (64)
36 (29)
12 (10)
28(54)
7 (13)
2 (4)
0.002
a
0.26
0.07
0.30
a
b
FM
n = 52
P value
1 (1)
76 (99)
2 (4)
50 (96)
0.57
44 (57)
33 (43)
28 (44)
24 (56)
0.72
a
13 (17)
64 (83)
7 (14)
45 (87)
0.81
4 (5)
73 (95)
2 (4)
50 (96)
1.0
13.0
140
12.0
140
0.23
4.0
130
4.0
120
0.33
7896%) in group FM (p = 0.76). In group WW with CM the threeyear OS was 91% (95% CI 8296% p = 0.41) and the three-year DSS
was 93% (95% CI 8597% p = 0.49).
Fig. 2 shows the Kaplan Meier survival curves of the mentioned
groups.
Total
200
PM
22
Discussion
Our current strategy in treating early stage OSCC is effective, as
it resulted in only 4.5% local recurrences, which concurs with the
lower limit of percentages mentioned in the literature ranging
from 4% to 22% [7,1416]. Most recurrences are located at the tongue (5.7%) followed by the cheek (4.5%) and oor of the mouth
(2.7%).
These results are similar to those reported by others [12,14,17].
Sixty-three percent of the OSCCs was resected with close margins.
In literature this percentage ranges between 15% and 42% [8,10].
An explanation could be our group distribution, which was based
on margin status. As our margins between 0 and 5 mm are
classied as close and others consider margins 61 mm or more
WW (with CM)
n = 77
CM
126
FM
52
RR
16
PORT
WW
1
RR
15
PORT
34
WW
77
RR
0
PORT
WW
52
Table 3
Adjuvant therapy in relation to patient groups and recurrences.
Variables (n = 200)
PM
n = 22
Recurrence
RR n = 31
PORT n = 39
WW n = 130
CM
n = 126
2
16
5
1
1
1
0
FM
n = 52
5
15
34
77
0
4
1
0
0
52
0
0
2
1 (3)
5 (13)
3 (2)
Bold: The amount of patients in relation to margin status and treatment modality. Italic: The amount of patients with local recurrence in relation to margin status and
treatment modality.
614
Figure 2. Kaplan Meier survival curves of group PM, CM, FM and group WW with CM.
as positive [4], this may have resulted in a shift of patients form the
positive margin to the close margin group. Another explanation is
our pathological analysis. Margins were measured in millimetres
with 1 decimal accuracy and were not rounded, which probably
led to a bigger proportion of OSCCs resected with close margins.
The management of positive or close margins is a recurring
point of discussion among clinicians. There is no consensus about
which adjuvant treatment modality to use in case of a positive
margin [14,9]. As positive margins have an adverse effect on local
control, many authors state adjuvant treatment is justied [1
4,9,12,18]. Our study underscores this statement. Even despite
adjuvant treatment, 9% of the resected OSCC with a PM recurred
locally compared with 4.0% in the CM group and 3.8% in the FM
group.
In case of close margins, no consensus about the necessity of
adjuvant treatment exists. Some authors state close margins between 0 and 5 mm are strongly related to local control
[3,4,12,18], while others refute this [5,7,19]. No prospective randomised clinical trials are available to answer this question. Most
studies have a retrospective design [8,11,12,20,21] and because
of the low numbers of local recurrences, signicant conclusions
are often impossible to draw [11,12]. We found a similar level of
recurrence, OS and DSS in the CM and FM groups, which suggests
that free margin status is irrelevant. However, the CM group is an
inhomogeneous group: 34/126 (27%) patients received PORT, 15/
126 (12%), underwent RR and 77/126 (61%) received no adjuvant
therapy at all (Table 3). Therefore, only a comparison between
the WW group with CM and FM groups is justied (Table 5),
showing a local recurrence of 1.3% and 3.8% respectively (not signicant). As, apart from resection margins averaging 3 mm in the
WW group and 6 mm in the FM group, no signicant differences
in pathological parameters, OS and DSS between these groups were
seen, it can be concluded that where local recurrence risk is concerned, at least a free margin of 3 mm is just as safe as one of
6 mm. Indeed, it could be argued that the whole concept of using
a free margin status is irrelevant in early oral cancer: once resection margins are clear, recurrence risk is extremely low, as has previously been demonstrated [5,7]. A substantial proportion of the
CM group may have undergone PORT or RR without evident necessity or benet while causing extra morbidity and expenses [22,23].
In comparison to our 19851994 cohort of Stage 12 oral cancers,
in which no adjuvant treatment for CM was given, this treatment
strategy did not alter the risk of local recurrence [7]. Therefore
the close margin concept introduced a decade ago seems irrelevant in making decisions on adjuvant treatment for Stage 12 oral
cancers [10,11].
Many authors suggest that histological parameters such as
depth of tumour inltration [14,18,2426], peri-neural ingrowth
vasoinvasive growth and spidery growth are also related to poor
local control [2,10,12,19,20,27], and hence indicate adjuvant treatment. Our study could not endorse that statement. Also in this
study, a multivariate analysis of pathological parameters was not
possible due to the small recurrence numbers.
In case of PM or CM, local adjuvant treatment is performed in a
considerably proportion of the cases. Some authors suggest RR
[2,3,8,14]. Others suggest postoperative radiotherapy [5,9,19,28],
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
615