87 / 220
late dysphagia. Sixty-two patients (74%) have been followed
beyond 5 years, or have experienced severe late dysphagia, death
or disease recurrence.
The actuarial 5-year rate of locoregional failure was 20% (95% CI
12–31%), distant metastases 16% (95% CI 9–27%) and overall sur-
vival 70% (95% CI 58–80%). Fifty-three of 84 patients (63%) have
either experienced toxicity, recurrence, death or have been fol-
lowed beyond 9 years. Ten patients underwent a salvage laryngec-
tomy after recurrence for a 5-year cumulative incidence of 15%
(95% CI 8–25%). Eight of the ten salvage laryngectomies were suc-
cessful and the patients were alive and without evidence of disease
at the time of their last follow-up.
Late toxicity
Twenty-two patients experienced severe late dysphagia. The
cumulative incidence of severe late dysphagia and its components
are presented in
Table 2
and
Fig. 1
. The overall cumulative inci-
dence of severe late dysphagia at 5 years was 26.5% (95% CI
15.2–37.8%). Sixty-eight percent of patients required feeding tube
support on-treatment (
Fig. 2
) but at one year after radiotherapy
only 1.8% remained feeding tube dependent (95% CI 0.2–11.2%).
No patient required a laryngectomy for toxicity. Stricture dilation
within the first year after radiotherapy was the most common sev-
ere late dysphagia event. Of the 18 patients who required stricture
dilation, 12 required multiple dilations. The median number of
dilations for those who underwent dilation was 2 (range 1–8). Of
the 22 patients who experienced severe late dysphagia, four
(18%) experienced their first event beyond 5 years.
Cause of death
At the time of last follow-up, 31 patients had died (37%).
Table 3
presents the distributions of deaths observed in comparison to the
distribution seen on RTOG 91-11. The index head and neck cancer
was the most common cause of death in both cohorts. Two patients
in our series expired as a result of neutropenic fever while on-
treatment. The cause of death could not be determined in 6
patients (19% deaths) who expired out of state or out of the coun-
try. The distribution is comparable to RTOG 91-11 (Pearson Chi-
square
p
= 0.454). Of note, the specific number of deaths observed
on RTOG 91-11 due to aspiration pneumonia is not specified, but
three late RTOG grade 5 toxicities were observed within the con-
current chemotherapy arm: one pharynx/esophagus, one larynx
and one ‘‘other” event, leading to a maximum crude rate of 2% or
less.
Factors associated with severe late dysphagia
To investigate clinical factors which may be associated with
severe late dysphagia, a Fine-Gray competing risk regression was
performed. Results are presented in
Table 4
. Among all patient,
tumor and treatment factors entered into the univariate regression,
twice-daily radiotherapy fractionation was the only statistically-
significant association with increased severe late dysphagia (HR
2.51, 95% CI 1.10–5.72,
p
= 0.028). The use of single agent cisplatin
rather than multiagent chemotherapy, or IMRT as opposed to 3D
planning were not associated with a reduction in severe late dys-
phagia on univariate analysis. A multivariate analysis was not per-
formed given the univariate results.
Discussion
In this study patients with larynx cancer who met the inclusion
criteria of RTOG 91-11 were retrospectively identified and a
detailed time-to-event analysis of severe late dysphagia was per-
formed while accounting for the competing risks of recurrence or
death. This is the first analysis to apply a competing risk analysis
Table 1
Study demographics (
n
= 84).
Age at first diagnosis
Median (Range)
60 (43–76)
Race
Caucasian
72 (87%)
African American
10 (12%)
Hispanic
1 (1%)
Gender
Male
60 (71%)
Female
24 (29%)
Smoking history
Never smoker
3 (4%)
Former smoker
(Quit >3 months)
38 (45%)
Current smoker
33 (39%)
Use during or after
radiation
9 (11%)
Unknown smoking
history
1 (1%)
Karnofsky score
Median (Range)
90 (80–90)
Tobacco pack-years
Median (Range)
40 (0–200)
Heavy alcohol consumption? No
68 (81%)
Yes
16 (19%)
Larynx subsite
Supraglottic
60 (71%)
Glottic
24 (29%)
T Classification
2
13 (16%)
3
65 (77%)
4
6 (7%)
N Classification
0
31 (37%)
1
18 (21%)
2a
3 (4%)
2b
12 (14%)
2c
19 (23%)
3
1 (1%)
Grouped stage
III
45 (54%)
IV
39 (46%)
Lymph node dissection
No
74 (88%)
Yes
10 (12%)
Chemotherapy
Cisplatin (CP)
17 (20%)
CP/5FU
63 (75%)
CP/5FU with Gefitinib 1 (1%)
Other multiagent
3 (4%)
Radiation type
3D-RT
63 (75%)
IMRT
21 (25%)
Altered fractionation
Daily
46 (55%)
BID
31 (37%)
6-Fractions per week 7 (8%)
Feeding tube placed during
treatment
No
27 (33%)
Yes
56 (68%)
Dose of RT
Median (Range)
72 Gy (62.4–
74.4 Gy)
Number of fractions
Median (Range)
36 (32–62)
Duration of RT (Days)
Median (Range)
46 (29–64)
Months of follow-up
(Survivors)
Median (Range)
53 (8.8–180)
Number of follow-up visits
Median (Range)
14 (0–44)
Table 2
Cumulative incidence (CI) of severe late dysphagia and its components (cumulative
incidence and 95% confidence intervals are listed). Twenty-two patients experienced
severe late dysphagia and 4 experienced the first event beyond 5 years.
CI of stricture dilation at 5 years
17.2% (8.9–25.6%)
CI of late feeding tube placed at 5 years
3.8% (0–8.0%)
CI of aspiration admission at 5 years
2.8% (0–6.9%)
Feeding tube dependent at 1 year
1.8% (0.2–11.2%)
CI of severe late dysphagia at 5 years
26.5% (15.2–37.8%)
M.C. Ward et al. / Oral Oncology 57 (2016) 21–26
67