2017 Section 7 Green Book

171 / 220

approach is preferable for long term swallow outcomes. We have
examined a retrospective cohort of patients who all received con-
current chemo-radiotherapy using whole-field parotid-sparing
IMRT for locally advanced oropharyngeal carcinoma, using the
MDADI as a validated patient-reported tool, with long term follow
up of at least 2 years post-treatment. These data suggest that the
use of a prophylactic gastrostomy results in statistically inferior
overall MDADI scores, as well as in the global, emotional and func-
tional subscales, with a small statistically non-significant benefit in
the physical domain.
The clinical significance of these results is dependent upon the
extent to which these two groups of patients are comparable. All
patients received concurrent chemotherapy and bilateral neck IMRT
and did not require therapeutic enteral feeding prior to treatment.
The matched pair analysis was performed to minimise differences
in swallow outcome which may have been due to tumour stage,
which are recognised to influence long term swallow function
[2,7]. The selection of feeding route was dependent upon clinician
and patient preference, and all patients were entered into routine
programmes of dietetic and speech and language therapy support
during and after treatment. There was no difference in baseline
swallow function between these two groups of patients, measured
using a simple dietetic consistency scale. In addition there were no
significant differences in patient demographics, tumour stage, treat-
ment details between the two groups. Despite this, it is not possible
to completely exclude the possibility that baseline factors may have
influenced the choice of approach to enteral feeding and conse-
quently confound possible associations with swallow function.
The T stage and N stage match involved grouping stages together
e.g. N0 and N1, to allow an adequate number of patients to be
matched for subsequent analysis. There was a higher number of
N0 patients within the prophylactic gastrostomy group, although
there was no significant difference between T and N stages, and N0
nodal stage might be expected to be associated with superior swal-
lowing outcomes. A slightly higher proportion of patients in the pro-
phylactic gastrostomy group received induction chemotherapy,
possibly reflecting a perceived clinical preference for using a gas-
trostomy to support patients through treatment involving induction
and concurrent chemotherapy.
There are some limitations to this study. We do not have human
papilloma virus (HPV) status available for this a useful proportion
of this historical cohort of patients as it was not being routinely
tested at our institution in this era. However, it seems likely that
HPV status is balanced across the two groups as the proportion
with current or previous smoking status was similar, as was the
proportion of patients with advanced nodal disease. There is no
data to suggest that there are differences in the impact of chemora-
diotherapy upon late dysphagia risk depending upon HPV status. It
should also be noted that the mean parotid doses achieved with
the compartmental outlining methods are considerably higher
than we would currently expect with current volumetric outlining
and more advanced IMRT delivery techniques; it is possible that
this may have impacted upon the overall swallow function.
Conclusion
Many factors may influence long term swallow recovery post-
chemoradiotherapy, including patient characteristics, baseline
swallow function, tumour factors, smoking status, and swallowing
support and rehabilitation provided during and after treatment
[2,10]. The timing and route of enteral feeding tube may be an
important factor. This is an area in which previous randomised tri-
als [36] have failed to adequately recruit, and institutional out-
comes are important to inform practice. This matched pair
analysis reinforces concern over the potential for a prophylactic
gastrostomy to negatively impact upon long term swallow
recovery.
Conflict of interest statement
We have no conflicts of interest.
References
[1]
.
[2]
.
[3]
.
[4]
[5]
[6]
.
[7]
.
[8]
[9]
.
[10]
[11]
.
[12]
.
[13]
[14]
.
[15]
[16]
.
[17]
[18]
.
[19]
[20]
[21]
[22]
B. Sethugavalar et al. / Oral Oncology 59 (2016) 80–85
151