Annual Meeting Proceedings Part 1 - American Society of Clinical ...

362s Head and Neck Cancer
5524 Poster Discussion Session (Board #14), Sat, 8:00 AM-12:00 PM and
12:00 PM-1:00 PM
New DNA methylation markers associated with oral cancer (OC) development
(dvlpt). Presenting Author: Pierre Saintigny, University of Texas M. D.
Anderson Cancer Center, Houston, TX
Background: We have reported that DNA methyltransferase-3B (DNMT3B)
gene expression is a strong predictor of OC dvlpt. Using high-throughput
DNA methylation profiles of oral preneoplastic lesions (OPL), we identified
86 candidate genes that were hypermethylated in patients (pts) developing
OC and suggested that a CpG island methylation phenotype may occur early
during OC dvlpt. Our aim was to validate some of the candidate genes and
to study the value of LINE1 repetitive element methylation, a surrogate of
global DNA methylation, as biomarkers of risk to develop OC. Methods:
Validation cohort included 40 pts with a median follow-up of 4.2 years,
including 14 pts who developed OC. None of them were used in the
discovery phase. Frozen OPL were collected prospectively and subject to
DNA extraction. We performed a quantitative analysis of the degree of
methylation of LINE1, and AGTR1, FOXI2, PENK, and HOXA9 promoters
CpG sites using pyrosequencing. Results: The degree of methylation of
AGTR1 (Mann-Whitney P�0.0001), FOXI2 (P�0.0002), PENK
(P�0.0001), but not HOXA9 (P�0.0714) was significantly higher in pts
who developed OC. On the contrary, LINE1 methylation was significantly
lower in pts who developed OC (P�0.0001). The median percent of
methylation was used to dichotomize the pts into high versus low methylation
levels. Oral cancer-free survival (OCFS) was significantly worse in pts
with high levels of AGTR1 (Log-rank P�0.0229), FOXI2 (P�0.0170), and
PENK (P�0.0142) promoter methylation. No significant difference was
found with HOXA9. A Methylation Index (MI) was developed by averaging
the percent of methylation of AGTR1, FOXI2, and PENK promoters; pts
with a high MI had a worse OCFS (P�0.0031). On the other hand, pts with
low levels of LINE1 methylation had a significantly worse OCFS (P�0.0118).
Finally, in 26 pts, a positive correlation was observed between DNMT3B
gene expression levels and the MI (rho�0.65, P�0.0003); surprisingly,
there was a negative correlation with LINE1 methylation (rho�-0.74,
P�0.0001). Conclusions: AGTR1, FOXI2 and PENK promoter methylation
may be associated with increased OC risk in pts with OPL and correlate with
DNMT3B expression. Global DNA hypomethylation is an early event in OC
dvlpt.
5526 Poster Discussion Session (Board #16), Sat, 8:00 AM-12:00 PM and
12:00 PM-1:00 PM
Effects of low-level laser therapy in the prevention and treatment of
concurrent chemoradiotherapy induced oral mucositis: A triple-blind randomized
controlled trial. Presenting Author: P.U. Prakash Saxena, Department
of Radiotherapy & Oncology, Kasturba Medical College, Udupi, India
Background: Oral mucositis (OM) is the most cumbersome acute side effect
of concurrent chemoradiotherapy (CCRT) for head and neck cancer (HNC).
OM associated pain affects oral function also may lead to CCRT break.
Several modalities have been tried to prevent and treat this complication
but none has been proved successful till date. We used prophylactic Low
Level Laser Therapy (LLLT) for the prevention and treatment of CCRT
induced OM. Methods: In this triple blind study, 221 HNC patients
scheduled to undergo CCRT were randomized into laser (n�111) and
placebo (n�110) group. Laser group received LLLT (Helium Neon laser,
��632.8nm, P�24m, W�3.0J/cm2 ) while placebo received sham treatment
daily prior to radiation for 45days. Oral Mucositis (OM) (RTOG/EORTC
Scale), oral pain (Visual Analog Scale - VAS), dysphagia (Functional
Impairment Scale - FIS), weight loss, duration of analgesic use were
assessed. Data was analyzed using frequencies and percentage, generalized
estimating equations (GEE) and Odds ratio. Results: There was
significant reduction in incidence of severe OM (F�16.64, degree of
freedom df�8,876, p�0.0001) and its associated pain (F�25.06,
df�8,876, p�0.0001) and dysphagia (F�20.17, df�8,876, p�0.0001)
and loss of weight (F�87.56, df�8,876, p�0.0001) when compared
between the Laser and placebo group. The duration of step III analgesia
required was also significantly lower (p�0.001) in laser (3.2 �1.4days)
than placebo (6.7 �2.6days) group. Conclusions: LLLT was able to
decrease the incidence of CCRT induced severe OM its associated pain,
dysphagia, weight loss and the duration of analgesic use.
Statistical analysis of OM and FIS scores between laser and placebo group.
Week
Laser group
OR (95% CI)
Oral mucositis Functional Impairment Scale
Placebo group
OR (95% CI)
Laser group
OR (95% CI)
Placebo group
OR (95% CI)
7 19.05 (9.78-37.12) 8.18 (3.03-22.05) 12 (6.77-21.2) 5.24 (3.15-8.72)
6 12.72 (6.73-24.04) 5.17 (1.83-14.58) 9.5 (5.45-16.53) 4.7 (2.85-7.74)
5 7.06 (3.88-12.83) 3.86 (1.37-10.84) 6.1 (3.62-10.28) 3.89 (2.35-6.45)
4 3.65 (2.05-6.49) 3.24 (1.27-8.27) 2.7 (1.74-4.2) 2.48 (1.57-3.91)
3 Ref Ref Ref Ref
5525 Poster Discussion Session (Board #15), Sat, 8:00 AM-12:00 PM and
12:00 PM-1:00 PM
Hearing evaluation of head and neck cancer patients (HNCP): Comparison
of adverse event (AE) criteria. Presenting Author: A. Dimitrios Colevas,
Stanford University School of Medicine, Stanford, CA
Background: The incidence and severity of baseline hearing abnormality in
HNCP is poorly characterized, as is the subsequent CDDP associated
hearing loss. While formal behavioral audiograms can characterize CDDP
associated hearing loss in great detail, most oncologists either rely on
subjective evaluation of hearing loss or grade hearing impairment using the
CTCAE. Two new grading systems for grading cisplatin ototoxicity have
been developed for use in pediatrics, the Brock and Chang systems. This
report is the first evaluation of these 3 hearing loss criteria in an adult
cancer patient population. Methods: We conducted an electronic medical
records (EMR) search for all HNCP between 2004 and 2010 treated at the
Stanford Cancer Institute (SCI) and identified which patients had audiograms
in the EMR, and who had received CDDP. Three investigators (RL,
ADC, KC) independently graded these audiograms using the CTCAE v3 and
v4, Brock and Chang criteria. Baseline distributions, changes following
CDDP dosing, and inter-observer variability were calculated for hearing
impairment grading using all four criteria. Results: We evaluated 460
audiograms in 111 patients. Because 282/460 of the audiograms did not
have threshold shifts measured at 1,2,3,4,6 and 8 kHz, CTCAE v4 could
not be applied to them . We therefore limited all further comparisons to
CTCv3, Brock and Chang criteria. The table below summarizes the analysis.
Conclusions: The Brock and Chang ototoxity criteria scales deliver a
distribution of hearing AE grades at baseline more compatible than the
CTCAE v3 with respect to the expectation that most HNCP do not have
severe baseline hearing loss. Both the Brock and Chang scales are more
sensitive than the CTCv3 to CDDP associated hearing changes. The Brock
scale, followed by the Chang scale, demonstrated highest inter-rater
agreement. Additional data concerning performance characteristics and
utility of these three grading systems will be presented. Replacement of the
present hearing impairment criteria in the CTCAE with either the Brock or
Chang scales should be considered.
Ototoxicity scale CTCAE v3 Brock Chang
Baseline median grade 2 0 0
% inter-rater agreement 69 91 80
% with grade changes following CDDP 33 44 47
5527 Poster Discussion Session (Board #18), Sat, 8:00 AM-12:00 PM and
12:00 PM-1:00 PM
ARIX: A randomized trial of acupuncture versus oral care sessions in
patients with chronic radiation-induced xerostomia following treatment for
head and neck cancer. Presenting Author: Richard Simcock, Sussex Cancer
Centre, Brighton, United Kingdom
Background: Radiation induced xerostomia impairs quality of life of in head
and neck cancer survivors. Prior small non-randomised studies suggest that
acupuncture may alleviate symptoms. We report the first large randomised
trial comparing 8 weeks of group acupuncture with oral care sessions as a
control. Methods: 145 patients with radiation induced xerostomia�18
months were recruited from 7 UK Cancer centres. Subjects received
standardised group sessions of oral care (OC) education and 8 sessions of
weekly group acupuncture (A) in a randomised crossover design with a 4
week washout (Group One OC/A and Group Two A/OC). Acupuncture was
standardised to a technique previously piloted (3 auricular points and two
distal points bilaterally). Patients completed the EORTC QLQ H&N questionnaire
at baseline and weeks 5, 9, 13, 17 and 21 in addition to rating 4 key
dry mouth symptoms. The primary outcome was patient reported improvement
in dry mouth symptoms. The statistical analysis was longitudinal
based on logistic regression models and adjustment for potential carry over.
Stimulated and unstimulated saliva was measured. Results: An improvement
for 5/7 symptoms was noted following acupuncture. In Group One
24% showed improvement 8 weeks following A (19% after OC). In Group
Two 26% reported improvement in severe dry mouth after A (14% after
OC). The estimated OR of improved dry mouth 5 weeks after A compared to
OC was 2.0 (p�0.031), after adjusting for effects of time, potential
carry-over, centre and patient characteristics. The estimated odds ratios of
improvement in severity of symptoms for A compared to OC were (OR�1.67,
p�0.048) for sticky saliva, (OR�2.08, p�0.011) for needing to sip to
swallow food and (OR�1.71, p�0.013) for waking up at night needing to
drink. Mean attendance rate for A was 89% and 80% for OC. There was no
significant change in saliva production over time or by intervention. Mean
global QoL score did not change significantly over time either within or
between groups. No adverse events were seen. Conclusions: Eight sessions
of weekly group acupuncture provides significantly better relief of radiation
induced xerostomia than group oral care education.
Visit abstract.asco.org and search by abstract for the full list of abstract authors and their disclosure information.