De-escalation of surgery for early oral cancer – is it oncologically safe?

DoCFa

b

c

d

e

AA

A

Twerui1atT

K

I

T(ao

pf

cb(a(

0

British Journal of Oral and Maxillofacial Surgery 51 (2013) 3036

Available online at www.sciencedirect.com

e-escalation of surgery for early oral cancer is itncologically safe?onor P. Barry a,, Chetan Katre b, Elena Papa c, James S. Brown a, Richard J. Shaw a,d,azilet Bekiroglu a, Derek Lowe a, Simon N. Rogers a,e

Regional Maxillofacial Unit, University Hospital Aintree, Liverpool L9 7LN, UKMaxillofacial Unit, North Manchester General Hospital, Manchester M85RB, UKMaxillofacial Unit, University Hospital Evagelismos, Athens, GreeceDepartment of Surgery and Oncology, School of Cancer Studies, University of Liverpool, Liverpool, UKEvidence-Based Practice Research Centre (EPRC), Faculty of Health, Edge Hill University, St Helens Road, Ormskirk L39 4QP, UK

ccepted 24 February 2012vailable online 22 March 2012

bstract

his study is a review of practice for patients with T1 or T2 squamous cell carcinoma (SCC) of the anterior tongue and floor of the mouthho presented to the regional maxillofacial unit in Liverpool between 1992 and 2007. We examined trends in management and analysed their

ffects on resection margins, recurrence, and survival. The Liverpool head and neck oncology database was used to identify patients, and toetrieve their clinical, surgical, and pathological data. When data were missing the case notes and pathology records were reviewed. Followp was taken to January 2011. A total of 382 patients were included. Despite more conservative treatment with closer resection margins (27%n 19921995 and 60% in 20042007), fewer free flaps (79% in 19921995 and 38% in 20042007), and less adjuvant radiotherapy (37% in9921995 and 22% in 20042007), there has been no significant increase in local recurrence (14% in 19921996 and 8% in 20042007),
nd overall survival has not been adversely affected. This is most striking when T1 tumours are considered in isolation with a consistent trendowards fewer clear margins (95% in 19921995 and 28% in 20042007) and fewer free flaps (53% in 19921995 and 11% in 20042007).he case mix was similar over the study period. These data support a more conservative approach to the management of early oral cancer.
2012 Published by Elsevier Ltd on behalf of The British Association of Oral and Maxillofacial Surgeons.

urgery

tmw

eywords: Survival; Recurrence; Pathological margins; Mouth neoplasm; S

ntroduction

he principal management for oral squamous cell carcinoma
SCC) with curative intent is radical primary operation, whichims to remove the entire tumour with resection marginsf 1.5 cm of normal tissue. This is followed by reconstruc-
Corresponding author at: Regional Maxillofacial Unit, University Hos-ital Aintree, Liverpool L9 7LN, UK. Tel.: +44 151 529 5287;ax: +44 151 529 5288.

E-mail addresses: [email protected] (C.P. Barry),[email protected] (C. Katre), [email protected] (E. Papa),[email protected] (J.S. Brown), [email protected]. Shaw), [email protected] (F. Bekiroglu),[email protected] (D. Lowe), [email protected]. Rogers).

pth

htafiohT

266-4356/$ see front matter 2012 Published by Elsevier Ltd on behalf of The Britisdoi:10.1016/j.bjoms.2012.02.014

; Radiotherapy; Oral cancer; Free flap

ion using free tissue transfer where indicated to promote theost functional outcome. Selective neck dissection is donehen invasion is estimated to exceed 4 mm in depth or torovide access to vessels in the neck to facilitate reconstruc-ion. Prescription for adjuvant radiotherapy is based on theistopathological findings of the resection specimen.

Various predictors of locoregional recurrence and survivalave been described,13 and the adverse prognostic implica-ions of compromised resection margins for local recurrencend survival have been well documented.35 Published datarom this unit suggest that local recurrence could be doubledf the margin is close (15 mm),3 but this is for all stages of
ral cancer and not just for T1 and T2. Some recent evidenceas called into question what constitutes a safe margin.69
here is however consensus about the prognostic importance

h Association of Oral and Maxillofacial Surgeons.
dx.doi.org/10.1016/j.bjoms.2012.02.014http://www.sciencedirect.com/science/journal/02664356mailto:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]/10.1016/j.bjoms.2012.02.014

C.P. Barry et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 3036 31

Table 1Patients characteristics at presentation. Data are number (%) unless otherwise stated.

All patients with T1 or T2 tumours 19921995 19961999 20002003 20042007 P-value*

(n = 63) (n = 89) (n = 99) (n = 131)

ClinicalT1 19 (30) 39 (44) 50 (51) 65 (50) 0.05T2 44 (70) 50 (56) 49 (49) 66 (50)

SexMale 40 (63) 59 (66) 63 (64) 82 (63) 0.96Mean (SD) age (years) 58 (11) 61 (13) 62 (13) 62 (13) 0.28

Tumour siteTongue (anterior 2/3) 32 (51) 46 (52) 53 (54) 82 (63) 0.27Floor of mouth 31 (49) 43 (48) 46 (46) 49 (37)Clinical N+ 12 (19) 12 (13) 21 (21) 25 (19) 0.57

Patients with T1 tumours (n = 19) (n = 39) (n = 50) (n = 65)


Tumour siteTongue (anterior 2/3) 9 (47) 20 (51) 29 (58) 44 (68) 0.25Floor of mouth 10 (53) 19 (49) 21 (42) 21 (32)Clinical N+ 1 (5) 1 (3) 4 (8) 5 (8) 0.71



Tumour siteTongue (anterior 2/3) 23 (52) 26 (52) 24 (49) 38 (58) 0.83Floor of mouth 21 (48) 24 (48) 25 (51) 28 (42)Clinical N+ 11 (25) 11 (22) 17 (35) 20 (30) 0.51 Chi-square test between the four time periods apart from one-way ANOVA for age.

ohar

pbattStittt

afms

en

P

SatawrbctcdS

f extracapsular spread (ECS) in cervical lymph nodes, whichas been shown to double the chances of local recurrencend distant metastasis, and triple the chances of regionalecurrence.10

While cure and survival are of primary concern to theatient and the surgeon, radical surgery causes serious mor-idity. Free flaps are necessary for the reconstruction of largend composite defects, and they have enabled large oralumours to be resected adequately without destructive func-ional and aesthetic morbidity. However, for early stage oralCC, free tissue transfer can be associated with poorer func-

ional outcome11 and health-related quality of life12,13 thanf primary closure, local flaps, or healing by secondary inten-ion were possible. Also, operations that last a day, admissiono the ITU, and prolonged hospital stay, are a large burden onhe healthcare budget.14,15

There is a clinical impression in our unit that we havedopted a more conservative approach to primary surgeryor early oral SCC of the anterior tongue and floor of the
outh (smaller resections, and fewer free flaps and neck dis-
ections). We aimed to examine this trend and to analyse thepit

ffects of a de-escalation of primary surgery on recurrence,eed for adjuvant radiotherapy, and survival.

atients and methods

ince 1992 all patients diagnosed with head and neck cancert the regional maxillofacial unit, University Hospital, Ain-ree, have had their details entered on to a computerised headnd neck database. We used the database to identify patientsith clinically staged T1 and T2 primary SCC of the ante-

ior tongue and floor of the mouth who presented to the unitetween January 1992 and December 2007. Clinical, surgi-al, and pathological details, and data on recurrence relatingo the primary tumour were retrieved from the database, andase notes and pathology records were reviewed for missingata. Mortality was tracked through the Office for Nationaltatistics (ONS) to 1 January 2011.

We divided the 16-year period (19922007) into 4-yeareriods to consider temporal trend. Results are presented andnterpreted primarily for all early tumours combined, but theables are split so that results for T1 and T2 stages can be

32 C.P. Barry et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 3036

Table 2Treatment and neck dissection. Data are number (%).

All patients with early T1 or T2 tumours 19921995 19961999 20002003 20042007 P-value*

(n = 63) (n = 89) (n = 99) (n = 131)

Operation alone 40 (63) 67 (75) 75 (76) 102 (78) 0.18Operation and adjuvant radiotherapy 23 (37) 22 (25) 24 (24) 29 (22)No free flap 13 (21) 32 (36) 53 (54) 81 (62)Soft flap 45 (71) 54 (61) 42 (42) 48 (37)

C.P. Barry et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 3036 33

Table 3Histological results. Data are number (%).

All patients with T1 or T2 tumours 19921995 19961999 20002003 20042007 P-value*

(n = 63) (n = 89) (n = 99) (n = 131)

Tumour differentiationPoor 7 (11) 6/86 (7) 2/86 (2) 4/113 (4)Moderate 36 (57) 51/86 (59) 52/86 (60) 66/113 (58) 0.82**

Well 20 (32) 29/86 (34) 32/86 (37) 43/113 (38)

MarginsClear 40 (63) 54 (61) 41/96 (43) 35/124 (28)

34 C.P. Barry et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 3036

Table 4Recurrence. Data are number (%).

All patients with early T1 or T2 tumours 1992995 1996999 2000003 2004007 P value*

(n = 63) (n = 89) (n = 99) (n = 131)

Local 9 (14) 2 (2) 5 (5) 10 (8) 0.03Regional 8 (13) 5 (6) 10 (10) 15 (11) 0.44Local or regional, or both 15 (24) 7 (8) 15 (15) 23 (18) 0.06Distant 2 (3) 4 (4) 5 (5) 10 (8) 0.56


Local 0 0 4 (8) 4 (6) 0.22Regional 0 3 (8) 3 (6) 7 (11) 0.44Local or regional, or both 0 3 (8) 7 (14) 10 (15) 0.23Distant 0 1 (3) 1 (2) 3 (5) 0.70


Local 9 (20) 2 (4) 1 (2) 6 (9) 0.008Regional 8 (18) 2 (4) 7 (14) 8 (12) 0.18Local or regional, or both 15 (34) 4 (8) 8 (16) 13 (20) 0.01Distant 2 (5) 3 (6) 4 (8) 7 (11) 0.64

solstgtb

t

Fe

tpt1(wi

Chi square test between the four time periods.

hows that a more conservative approach to surgery for earlyral cancer is safe oncologically. To our knowledge this is theargest series to date to look specifically at outcomes in earlytage disease. The findings are strengthened by the consecu-ive nature and size of the group with detailed prospectivelyathered histological and outcome data. We accept the limi-ations inherent in any retrospective analysis so data should
e considered with caution.
The inclusion criteria were based on clinical ratherhan pathological stage because clinical stage is available

ig. 2. KaplanMeier survival curves by time period for patients withoutxtracapsular spread.

p(wrfs2

cgTirrmTwqo

rgtfeui

o the surgeon when planning treatment. Inevitably, theathological findings will cause a proportion of patientso be upstaged. Disease seemed to be more severe in the9921995 group with a higher percentage of T2 tumoursTable 1). There was also a higher incidence of ECS (Table 3),hich was by far the strongest predictor of survival. The

ncreased severity of disease may reflect a change in referralattern, as before the advent of the multidisciplinary teamTumour Board) in 2003, many patients with early cancersere still being treated in local hospitals. It may also

eflect an improvement in the awareness of oral cancer, andaster referral.16 From 1996 to 2007, the three groups wereimilar in terms of age and severity of disease, although the0042007 group had a second peak in incidence of ECS.

Overall, we have found a definite trend towards a moreonservative surgical approach with closer resection mar-ins, fewer neck dissections, and fewer free flaps (Table 2).he increase in close margins is associated with a reduction

n clear margins, but the incidence of involved margins hasemained constant at about 10%, which is within the reportedange for resections of oral SCC.3,6,9 The de-escalation isost striking when T1 tumours are considered in isolation.his shift in treatment reflects an increased awareness thathile local control should not be compromised by inade-uate resection, unnecessary functional morbidity caused byver resection should also be avoided.

As expected, outcome for the 19921995 group in terms ofecurrence and overall survival was poorer than for the otherroups (Table 4; Fig. 1). Overall survival has improved overhe study period. A slight dip in 20042007 may be accountedor by the increase in ECS in this group, but when we
xcluded ECS from the analysis, the trend reverted to a grad-al improvement (Fig. 2). These results mirror improvementsn survival already reported from this unit4 and others.17

al and M

Hdiwi

snpsaarsctiTmoi

spwmctpmmarrpI1rseior5

rfnmmpTtrwd

Httn

tTsirob

avsltatrrdi

R

1

C.P. Barry et al. / British Journal of Or

istological findings resulted in 13% of patients with T2isease being upstaged to pT3 or pT4, and survival was sim-lar to that previously reported for T3 and T4 oral SCC,2

hich suggests that they were not disadvantaged by theirnitial understaging.

Despite the increase in close margins, there has been noignificant increase in local recurrence in this study, and ofote, there has been no increase in radiotherapy to com-ensate for less aggressive surgery. However, there was alight upward trend in recurrence. While this may be partlyccounted for by increased ECS, it also serves as a warninggainst treatment being too conservative. Our local recur-ence rates compare well with those from other publishederies,68 but accurate comparison is difficult as few papersonfine their results to T1 and T2 disease of the anteriorongue and floor of the mouth. There is a lack of consensusnternationally as to what constitutes an adequate margin.he UK guidelines (which we follow) record the status ofucosal and deep margins, and designate margins of 5 mm

r more as clear, 15 mm as close, and less than 1 mm asnvolved.

The division between clear and close margins at 5 mmeems to be arbitrary. Nason et al.6 reported recurrence inatients with margins of 34 mm that was identical to thoseith margins of 5 mm or more. However, their data do notake clear the extent to which postoperative radiotherapy had

onfounded the results. Brandwein-Gensler et al.7 reportedhat status of resection margins was not an independentredictor of local recurrence in patients having similar treat-ents. Weijers et al.8 compared non-involved deep surgicalargins of less than 5 mm with those of more than 5 mm

nd found no significant difference in risk of local recur-ence. However, Sutton et al.3 and Rogers et al.4 who botheported data from this unit, described a significantly adverserognostic outcome associated with close margins (15 mm).t is possible that the inclusion of patients with margins of2 mm and those with margins of 35 mm has biased theiresults in this regard, and that re-examination of the data mayhow results that concur more closely with those of Nasont al.6 and Brandwein-Gensler et al.7 Al Rajhi et al.5 exam-ned the significance of resection margins specifically for T1r T2 disease of the anterior tongue and reported much lowerecurrence-free survival for patients with margins of less than

mm, but this included tumours with involved margins.The slight reduction in the rate of neck dissections partly

eflects a lower demand for access to vessels in the neckor microvascular surgery. Our unit still advocates electiveeck dissections for tumours that have invaded to an esti-ated depth of more than 4 mm18 and when the risk of nodaletastasis is thought to exceed 20%.19 In this group, 70% of

atients (45% with T1 and 91% with T2) had neck dissection.he role of elective neck dissection remains controversial in

his group and is being addressed by the SEND trial (theole of selective neck dissection used electively in patientsith early oral squamous cell carcinoma and no clinical evi-ence of lymph node metastasis in the neck: NCT00571883).

1

axillofacial Surgery 51 (2013) 3036 35

owever, the finding that 10% of patients with T1 or T2umours and a clinically N0 neck had ECS reinforces the facthat early stage disease can be aggressive, and emphasises theeed for a reliable preoperative method to screen the neck.

The avoidance of free flaps where possible results in a bet-er quality of life,12,13 and in better functional outcomes.11

he duration of operation and hospital stay,15 and con-equently chances for postoperative medical complicationsn this population, many of whom are elderly, are greatlyeduced with primary closure or local reconstruction. Notnly is this beneficial to the patient but it also reduces theurden on healthcare resources.14

Treatment in our unit for early SCC of the anterior tonguend floor of the mouth has shifted towards more conser-ative surgery with closer margins and fewer free flaps. Inelected cases, these tumours can be treated safely by wideocal excision with smaller margins than has previously beenhought. However, the slight increase in involved marginsnd recurrence, and the levelling of the survival curve forhe 20042007 group may serve as a warning that we haveeached the safe limits of de-escalation. Further work isequired to quantify what is a safe resection margin, and toiscover any improvement in function or quality of life thats associated with more conservative operations.

eferences

1. Huang TY, Hsu LP, Wen YH, Huang TT, Chou YF, Lee CF, et al. Pre-dictors of locoregional recurrence in early stage oral cavity cancer withfree surgical margins. Oral Oncol 2010;46:4955.

2. Woolgar J. Histopathological prognosticators in oral and oropharyngealsquamous cell carcinoma. Oral Oncol 2006;42:22939.

3. Sutton DN, Brown JS, Rogers SN, Vaughan ED, Woolgar JA. Theprognostic implications of the surgical margin in oral squamous cellcarcinoma. Int J Oral Maxillofac Surg 2003;32:304.

4. Rogers SN, Brown JS, Woolgar JA, Lowe D, Magennis P, Shaw RJ,et al. Survival following primary surgery for oral cancer. Oral Oncol2009;45:20111.

5. Al Rajhi N, Khafaga Y, El-Husseiny J, Saleem M, Mourad W,Al-Otieschan A, et al. Early stage carcinoma of oral tongue: prognosticfactors for local control and survival. Oral Oncol 2000;36:50814.

6. Nason RW, Binahmed A, Pathak KA, Abdoh AA, Sndor GK. What isthe adequate margin of surgical resection in oral cancer. Oral Surg OralMed Oral Pathol Oral Radiol Endod 2009;107:6259.

7. Brandwein-Gensler M, Teixeira MS, Lewis CM, Lee B, Rolnitzky L,Hille JJ, et al. Oral squamous cell carcinoma: histologic risk assessment,but not margin status, is strongly predictive of local disease-free andoverall survival. Am J Surg Pathol 2005;29:16778.

8. Weijers M, Snow GB, Bezemer DP, van dr Wal JE, van der Waal I. Thestatus of the deep surgical margins in tongue and floor of mouth squamouscell carcinoma and risk of local recurrence; an analysis of 68 patients.Int J Oral Maxillofac Surg 2004;33:1469.

9. Kurita H, Nakanishi Y, Nishizawa R, Yiao T, Kamata T, Koike T,Evans C, et al. Impact of different surgical margin conditions on localrecurrence of oral squamous cell carcinoma. Oral Oncol 2010;46:8147.

0. Shaw RJ, Lowe D, Woolgar JA, et al. Extracapsular spread in oral squa-
mous cell carcinoma. Head Neck 2010;32:71422.
1. Rogers SN, Lowe D, Fisher SE, Brown JS, Vaughan ED. Health-relatedquality of life and clinical function after primary surgery for oral cancer.Br J Oral Maxillofac Surg 2002;40:118.

3 al and M

1

1

1

1

1

1

1

6 C.P. Barry et al. / British Journal of Or

2. Rogers SN, Scott J, Chakrabati A, Lowe D. The patients account ofoutcome following primary surgery for oral and oropharyngeal can-cer using a quality of life questionnaire. Eur J Cancer Care (Engl)2008;17:1828.

3. Rogers SN, Lowe D, Yueh B, Weymuller Jr EA. The physical functionand social-emotion function subscales of the University of Washing-ton Quality of Life Questionnaire. Arch Otolaryngol Head Neck Surg2010;136:3527.

4. McCrory AL, Magnuson JS. Free tissue transfer versus pedicled flap inhead and neck reconstruction. Laryngoscope 2002;112:21615.

5. Rogers SN, Lowe D, Brown JS, Vaughan ED. The relationship betweenlength of stay and health-related quality of life in patients treated by

1

axillofacial Surgery 51 (2013) 3036

primary surgery for oral and oropharyngeal cancer. Int J Oral MaxillofacSurg 2001;30:20915.

6. Rogers SN, Hunter R, Lowe D. Awareness of oral cancer in the Merseyregion. Br J Oral Maxillofac Surg 2011;49:17681.

7. Shah JP, JohnsonN.W. Batsakis JG. Oral cancer. London: Martin Dunitz;2003.

8. OBrien CJ, Lauer CS, Fredricks S, Clifford AR, McNeill EB, Bagia JS,et al. Tumour thickness influences prognosis of T1 and T2 oral cavity
cancer but what thickness. Head Neck 2003;25:93745.
9. Weiss MH, Harrison LB, Isaacs RS. Use of decision analysis in planninga management strategy for the stage N0 neck. Arch Otolaryngol HeadNeck Surg 1994;120:699702.
De-escalation of surgery for early oral cancer is it oncologically safe?IntroductionPatients and methodsResultsDiscussionReferences

Documents

De-escalation of surgery for early oral cancer – is it oncologically safe?