Trial Information

Elective Versus Therapeutic Neck Dissection in the Treatment of Early Node Negative Squamous Cell Carcinoma of the Oral Cavity

ELECTIVE VERSUS THERAPEUTIC NECK DISSECTION IN THE TREATMENT OF EARLY NODE NEGATIVE SQUAMOUS
CELL CARCINOMA OF THE ORAL CAVITY

INTRODUCTION AND BACKGROUND INFORMATION:

Cervical metastasis is the single most important prognostic factor in head and neck cancers.
Appropriate management of the neck is therefore of paramount importance in the treatment of
these cancers. Should the neck be electively treated or there be a wait and watch policy?

Current practice is that the neck is always addressed whenever there is an increased
propensity to cervical metastasis or when patient follow-up is unreliable. This is usually
the case with large T staged cancers (T3/T4) and also when the neck is entered at the time
of surgery for the primary.

The situation is however different with regards early cancers of the oral cavity (T1/T2).
These cancers are usually treated with surgery where excision is through the per-oral route.
Elective neck dissection in such a situation is an additional surgical procedure with its
associated costs, prolonged hospitalization , morbidity and may be unnecessary in as high as
80% of patients who finally turn out to be pathologically node negative. There is
considerable debate whether to treat the neck electively the world over and there is no
definite evidence to support either the policy of wait and watch or elective neck treatment.
Proponents of elective neck treatment cite the fact that in those patients whose necks are
observed, nodal recurrences are detected at an advanced stage which is detrimental to the
ultimate outcome of the patient.

The major draw back of all trials apart from the retrospective nature is the fact that the
total numbers of patients in majority of them are very small (none more than 150, the
majority less than 100) and over different time spans to draw meaningful conclusions. It
would therefore be attractive to prove this point conclusively in a prospective randomized
trial.

To date there have been three prospective trials conducted to address this issue, two were
inconclusive while one showed a benefit towards elective neck dissection. The first and most
widely cited study was from the Institute Gustave- Roussy in 75 patients with oral cavity
cancers who had either elective or salvage RND.In this study, both groups of patients had
similar survival rates . Fakih et.al. ,reported similar findings in a study from our
institution where 100 patients were randomized into two groups, elective neck dissection
versus wait and watch. Only 70 of these were evaluable. Disease free survival (median
follow-up 20 months) was 52 % versus 63% in patients who underwent hemiglossectomy alone and
those who underwent hemiglossectomy and radical neck dissection, respectively. This
difference though slightly in favor of the elective neck dissection was not statistically
significant . Finally a prospective trial conducted by Kligerman and co-workers in Brazil
randomized 67 patients with T1-2 N0 SCC of the floor of mouth or oral tongue to receive
resection of the primary alone or resection and elective neck dissection. Occult disease was
seen in 21% of those who had received END and recurrence was seen in 24% of patients in this
group. By comparison the recurrence rate was 42% in the resection only group. This
translated into a 3.5-year disease free survival group of 49% in the resection only arm
versus 72% in the arm that received END (7). It is obvious from the above that the major
shortcoming in all three studies as in the retrospective studies was the relatively small
number of patients recruited in these trials (75, 100 & 63).

Thus there is clearly a need for a large randomized trial that will resolve the issue
either way once and for all. The origin of such a trial from the West is difficult due to
the relative lack of such cancers in that part of the world. At the Tata hospital itself on
conservative estimate 100-150 patients would be eligible for enrollment each year on such a
trial. Moreover the trial design and treatment arms are simple and this is ideal for a
multicentre trial in India as standardization and uniformity would be easily achievable
across all participating centers.

Presently there is no global consensus on the management of early T1/T2 cancers with a
clinically negative neck. As a consequence current practice the world over includes:

1. Elective neck dissection as a staging procedure . Clinical assessment of the lesion
with respect to size and depth of infiltration.

2. Assessment of the neck with ultrasound or ultrasound guided FNAC or CT / MRI.

3. Wait and watch policy (in the belief that neck dissection is unnecessary in up to 80%
of patients)

4. Interval neck dissection

5. Sentinel node biopsy. The trial proposed will help answer all the above questions
except the last.

AIMS OF STUDY

Primary Objective:

To demonstrate whether elective neck dissection (END) is equal or superior to the wait and
watch policy i.e.therapeutic neck dissection (TND) in the management of the clinically No
neck in early T1 /T2 cancers of the oral cavity.

Secondary Objective:

1. Does Ultrasound examination have any role in the routine initial workup of a node
negative patient?

2. How are patients ideally followed up -does sonography have a role or is clinical
examination sufficient.

3. Is assessment of tumor thickness by the surgeon at the time of initial surgery accurate
-Is there a correlation between tumor thickness as measured grossly by the surgeon, at
frozen section versus final histopathology.

4. Identify histological prognostic factors in the primary that may help identify a
sub-set of patients at an increased risk of cervical metastasis.

Stratification criteria:

1. Size

2. Sex

3. Site

4. Sonography

Ethical Concerns:

1. Should patients with ultrasonography findings suggestive of metastasis be removed from
randomization

- Ultrasonography is a simple and often used method to detect neck metastasis.
Sensitivity ranges from 50-75%. These figures are much higher in centers that routinely
practice sonography to detect neck metastasis as in some centers in Europe. However it
is not an accepted standard of care the world over including the USA and in India in
the initial management of a patient with a node negative neck. Moreover an accurate
ultrasound of the neck is extremely operator dependent and whether its routine use can
be recommended particularly in a country may be difficult.

- Specificity of ultrasonography in the clinically node negative neck is only 50-70%.
This can be increased by the addition of fine needle aspiration cytology of enlarged
nodes. Doing this procedure routinely in all patients will be difficult or next to
impossible given the expertise required both by the sonologist as well as the
pathologist particularly in our country.

- Costs- Even though the ultrasound is relatively inexpensive there are significant costs
and waiting periods involved given the number of patients with oral cancers in the
country.

- There is no evidence in literature that sono detection of occult metastasis and
immediate surgery translates into better control or survival.

- Interval neck dissection was performed in Vandenbrooks prospective series where the
primary was treated with radiotherapy and patients randomized to either neck dissection
or wait and watch six weeks later.

Protocol amendment details:

Protocol amendment was accepted by the Scientific ethics committee/Instititional Review
board on 29/12/2008 as follows:

All patients will undergo a pre randomization ultrasonography of neck. If the USG report
suggests metastasis patients would be treated according to merit to avoid ethical concerns
and if the USG report is normal or indeterminate, patients would be randomized for the
trial. This protocol amendment is done in view of the following 2 reasons.

Very often patients are uncomfortable to give consent when they are explained that the
initial ultrasonography report will be blinded and the findings will not be considered in
deciding the patient's treatment plan. This resulted in lower recruitment of patients and it
also raised ethical issues when the blind was not adequately maintained especially if the
sonography suggested metastasis.

To overcome a difference of opinion amongst examining clinicians about a clinically
significant node.

However, breaking the blind would have no bearing on the outcome / impact of this study.

Randomization (I): All patients will be randomly allocated into one of the two arms: wait
and watch policy group or elective neck dissection group. Both the arms will have similar
wide excision of the primary tumor in the oral cavity per oral route.

Randomization (II): Following surgery and after complete recovery prior to discharge,
patients will be randomized a second time for follow-up into two arms namely clinical
examination versus clinical examination and ultrasonography of the neck.

SURGICAL PROCEDURE:

Primary: The tumor will be excised after proper exposure via the per-oral route.

- Emphasis will be to achieve a wide clearance with tumor free margins to obviate the
need of subsequent radiotherapy to the primary, which would otherwise act as a
confounding factor.

Neck: Patients randomized to neck dissection will undergo a standard supra-omohyoid neck
dissection that will involve clearance of nodal Levels I, II, and III.Completion MND will be
done as and when required. Patients who develop metastatic adenopathy on follow-up will
undergo a modified neck dissection/radical neck dissection depending on the size of the
metastatic disease. Those who have nodes suggestive of metastasis on follow-up sonography,
will undergo a supraomohyoid dissection, frozen section followed by a modified neck
dissection if positive.

The depth of tumor infiltration though probably the most important individual prognostic
factor in deciding the likelihood of cervical metastasis, is unfortunately only available to
the clinician with the final histopathology report. Had this parameter been present at the
time of surgery the patient with an increased likelihood of metastasis could have undergone
an elective neck dissection while those with a lower incidence could have been saved an
unnecessary operation .This study would help find out the accuracy of correlation between
the gross assessments of thickness by the surgeon, on frozen section with the final
thickness on histopathology and would be of importance in a country like ours where oral
cancers are very common and the facilities of frozen as well as the expertise to measure
accurately tumor thickness at histopathology may be unavailable.

POST OPERATIVE RADIOTHERAPY:

The standard accepted criteria will be followed. Patients receiving post operative
radiotherapy for the primary will not be included in the main analysis but will be analyzed
separately (Escape clause).

END POINT OF STUDY:

The patients will be followed for a minimum of 24 months but an attempt will be made up to
60 months. The end point will be regional recurrence with a controlled primary or disease
free survival at 24 months, whichever is earlier. Local failures, distant metastases and
second primary will be documented. An attempt will be made to follow-up all patients till
death / completion of a minimum of 24 months. (This assumption is based on the fact that 90%
of recurrences in head & neck cancer occur within the first 24 months).

DATA COLLECTION, QUALITY CONTROL & ANALYSIS:

All data will be collected and kept at the Trial Office in the Clinical Research
Secretariat. Data will be collected on separate protocol sheets (clinical, radiological, and
pathological) and kept in the trial office along with the source documents. Trial
coordinator will be responsible for the upkeep of the data.

Assuming the baseline overall survival of 60%, for expected improvement in treatment arm of
10% the no. of patients required will be is 710 (355 in each arm) with α=0.05 (one sided) &
power of 80% (β=20). An interim analysis is planned at 250 events (death) occur.

Ultrasound in the routine follow-up of all patients will be labor intensive. However the
median follow-up to recurrence in all studies both retrospective/prospective has been 9
months on an average (range 6 months- 13months). It will therefore be important to follow
patients every vigilantly in the first 12 months from primary treatment. Follow-up schedules
will be: First visit: 4 weeks;First 6 months: 4-6 weeks; 6-12 months: 6-8 weeks; 12 months-
2 years: 8-12 weekly; Thereafter: 3 monthly.

Patients will be encouraged and counseled to come for check up on the earlier date within
the range allowed. All patients will be followed up by one of the investigators and entry
made both in the source document as well as the central registration cell at the CRS with
the study coordinator. At each examination patients will undergo a through head and neck
examination as well as an ultrasound if randomized to that arm.