14 / 79
6th ICHNO
page 15
6
th
ICHNO Conference
International Conference on innovative approaches in Head and Neck Oncology
16 – 18 March 2017
Barcelona, Spain
__________________________________________________________________________________________
OC-022 Association of patient derived xenograft
formation with oral cavity squamous cell cancer
outcomes
A. Hope
1
, C. Karamboulas
2
, W. Xu
3
, S. Huang
1
, J. Kim
1
, S.
Bratman
1
, J. Cho
1
, J. Ringash
1
, M. Giuliani
1
, A. Bayley
1
, J.
Waldron
1
, B. Perez-Ordonez
4
, D. Goldstein
5
, J. De
Almeida
5
, D. Brown
5
, J. Irish
5
, P. Gullane
5
, R. Gilbert
5
, B.
O'Sullivan
1
, L. Ailles
2
1
Princess Margaret Cancer Center/University of Toronto,
Radiation Medicine Program/Radiation Oncology,
Toronto, Canada
2
Ontario Cancer Institute, Stem Cell and Developmental
Biology, Toronto, Canada
3
Princess Margaret Cancer Center, Biostatistics, Toronto,
Canada
4
Toronto General Hospital/University of Toronto,
Pathology, Toronto, Canada
5
Princess Margaret Cancer Center/University of Toronto,
Otolaryngology - Head/Neck Surgery, Toronto, Canada
Purpose or Objective
To assess correlation between patient derived xenograft
formation (PDXF) and clinical outcomes following curative
treatment of oral cavity squamous cell carcinoma
(OCSCC).
Material and Methods
Patients undergoing curative surgery for OCSCC had tumor
samples extracted and implanted into NSG (Jackson) mice
to assess PDXF. Ten tumor samples per patient were
implanted in murine flank and time to PDXF from any of
the ten samples was recorded. Clinical outcomes for
patients were collected prospectively and charts reviewed
to confirm patient factors, pathologic details of surgery,
adjuvant therapies, patient survival, and tumor outcomes.
Univariable and multivariable analyses were performed to
determine correlations between PDXF and cancer
outcomes and overall survival.
Results
Between 2007-2015, 243 OCSCC patients had tumor
samples explanted to attempt PDXF. Of these, 161
samples demonstrated PDXF, with a median time to PDXF
of 50 days. Patients demonstrating PDXF had a high
frequency of advanced nodal stage (p<0.01), close margins
(p<0.03), and were more likely to receive adjuvant
therapy (p<0.02). PDXF+ patients had significantly
reduced 5-year overall survival (OS) (47% vs. 65%), higher
rate of distant metastases (DM) (22% vs. 6%), and a trend
to lower locoregional control (64% vs. 76%). OS was lower
for PDXF+ patients in groups treated with surgery alone
(64% vs. 88%) or with surgery and adjuvant radiation or
radiochemotherapy (52% vs 72%). Patients who
demonstrated PDXF within 8 weeks of surgery had lower
survival (60% vs 92%, HR: 3.0, p=0.01) and lower
locoregional control (70% vs. 90%, HR: 2.43, p=0.02) (see
Figure). DM rates were similar for all PDXF patients
regardless of time to PDXF, but was significantly higher
than for patients who never demonstrated PDXF.
Multivariable models of overall survival showed PDXF and
nodal status (N0 vs N+) as independently significant
(p<0.01).
Conclusion
PDXF in patients with OCSCC correlates with poor
oncologic outcomes and lower overall survival. PDXF may
provide a rapid (<8 week) biomarker to help select
patients for the most appropriate adjuvant therapy
following definitive surgery. PDXF in patients with OCSCC
should be assessed prospectively to determine if this
approach is feasible in a multi-institutional setting.
Symposium: New developments in surgery
SP-023 New developments in sentinel node biopsy of
head and neck cancer
S. Stoeckli
1
1
Kantonsspital St. Gallen, Other, St Gallen, Switzerland
Abstract text
Sentinel Node Biopsy (SNB) was introduced in the field of
head and neck surgical oncology more than fifteen years
ago. Meanwhile, the technique has been adopted for the
treatment of early oral squamous cell carcinomas in the
NCCN and several national guidelines. The feasibility,
safety and efficacy has been proven in many published
reports and meta-analyses. Several technological
developments have considerably improved the process of
lymphatic mapping, which consists of reliable detection
and safe excision of the lymph nodes at risk and their
thorough histopathologic work-up. The lecture reviews the
current available evidence on SNB, reflects the most
recent large scale studies, and gives an overview on
current challenges and future developments. In
particular, the role of new tracers, technological tools for
tracer detection and possibilities of intraoperative real-
time assessment of the seninel nodes are assessed.
SP-024 Integrated 3D virtual visualization of pathology
and reconstructive planning in head & neck cancer
M. Witjes
1
1
UMCG University Medical Center Groningen, Head and
neck, Groningen, The Netherlands
Abstract text
Introduction: 3D surgical planning software does not allow
tumor margin visualization which makes it difficult where
to plan the cutting planes for the mandibulectomy or
maxillectomy. 3D surgical planning is typically based on
CT imaging which does not allow adequate tumor
delineation. MRI allows for more precise tumor
delineation, but is not easily integrated in 3D surgical
planning due to limited bone segmentation options. We
therefore studied a new strategy based on fusion of CT and
MRI imaging in which MRI is used for tumor delineation and
CT for planning of the bone cutting planes.
Methods: MRI images were projected onto the CT images
for data fusion, which is typically supported by
radiotherapeutic planning software (Mirada, Mirada
Medical). Delineation of the gross tumour volume (GTV) on
MRI was performed using a semi-automated brush tool in
the software. The CT dataset, supplemented with the MRI-
based tumor delineation data, was exported as a DICOM
file and a radiotherapeutic structure set (RTSS) file.
Converting this data towards the surgical planning
software (Proplan, Materialise) required a conversion. A
compatibility algorithm was written using Matlab
(Mathworks). In the surgical software the cutting planes
were planned, utilizing the 3D visualized tumor. Cutting
guides were designed, 3D printed and sterilized for use in
the OR.
Results: Twenty patients were included after being
treated with either maxillectomy or mandibulectomy. On