Assessment of swallowing dysfunction using fees (flexible endoscopic evaluation of swallowing) in head and neck cancer patients undergoing radiotherapy

INTRODUCTION

With an annual toll of 1.1 million, head and neck cancer ranks as the seventh most prevalent cancer worldwide.^[1] Standard treatments such as radiotherapy (RT) and concurrent chemoradiotherapy (CCRT), vital for locally advanced cases, often introduce debilitating side effects, notably dysphagia, significantly impacting patients’ quality of life (QOL).^[2,3] In the challenging landscape of inoperable head and neck cancer, CCRT emerges as the primary therapeutic approach, albeit accompanied by the discordant notes of high toxicity, particularly affecting critical swallowing structures and resulting in dysphagia.^[4]

Dysphagia, stemming from both cancer and treatments, induces intricate anatomical and functional alterations, significantly diminishing QOL. Swallowing impairment, often underestimated, emerges as a post-RT concern, with silent aspiration posing threats to weakened patients.^[5]

For those traversing the realms of RT or CCRT, the journey involves an elevated risk of dysphagia, a severe side effect casting its shadow both in early and late stages.^[5] Prevalent in 40% of cases post-RT, dysphagia unveils life-threatening risks, including aspiration, encompassing anatomical and functional alterations that complicate swallowing, leading to reduced food intake.^[6] The often-overlooked consequence of swallowing impairment can culminate in silent aspiration, underscoring the imperative for objective assessments.^[7]

Objective assessments are imperative for posttreatment dysphagia evaluation. Techniques such as video fluoroscopic swallow study (VFSS) and flexible endoscopic evaluation of swallowing (FEES) offer nuanced insights.^[8] VFSS, a modified barium swallowing examination, provides real-time visualization, aiding in identifying aspiration causes and effects of different bolus volumes and textures.^[9] FEES, a preferred method for its ease, tolerability, bedside examination, and cost-effectiveness, allows direct visualization of the swallowing process.^[10]

Our study takes a comprehensive approach, gauging swallowing subjectively via the eating assessment tool-10 (EAT-10) score and objectively through FEES. The EAT-10, a validated tool, captures self-reported difficulties, while FEES facilitates real-time observation, delving into structural and functional aspects. This dual-pronged approach seeks a nuanced understanding of how dysphagia resonates in the realm of swallowing function, offering valuable insights into its profound impact.^[7]

OBJECTIVES

MATERIAL AND METHODS

A prospective observational study was conducted at our hospital in the Deglutition and Swallowing Suite from June 2022 to May 2023. Approval was obtained from the Scientific Research and Ethical Committee of our hospital. Informed consent was obtained from the participants before their involvement in the study.

Adults with confirmed squamous cell carcinoma or adenocarcinoma in the head and neck region, cancers where the treatment portal includes the pharyngeal region, aged 18 and above, were eligible. Inclusion required nonmetastatic disease, eastern cooperative oncology group (ECOG) performance status of I or II, and informed consent. Exclusions applied to cases with disease recurrence, metastasis, prior head and neck irradiation, or radiotherapy (RT) portals excluding the pharyngeal region. Patients with tracheostomy or Ryles tube were ineligible.

Collected demographic and baseline data at entry: age, sex, primary cancer site, tumour, node, metastasis (TNM) classification, total radiation dose, RT technique, and chemotherapy regimen. Preradiation assessment included eating assessment tool-10 questionnaire and objective analysis of penetration aspiration scale and bolus residue scale via FEES procedure. Simulation and planning computed tomography was performed for radiation therapy to contour, plan, and evaluate the RT treatment. Post-RT assessments at one and three months were done using EAT-10, PAS, and BRS via FEES, noting the time for scores to normalize. Data were compiled and analyzed in excel. Statistical analysis was conducted for interpretation and conclusions.

Tools used for analysis

EAT-10 (Eating assessment tool) questionnaire is used for subjective analysis.^[11] FEES procedure was carried out by SLP technician in our hospital in the deglutition and swallowing suite.

In the FEES procedure, equipment was prepared, and the patient’s consent was obtained. After applying local anesthesia, a lubricated endoscope was inserted through the nostril for real-time visualization of the pharynx, larynx, and upper esophageal sphincter. Swallowing trials with various consistencies were conducted, and the entire procedure was recorded [Figures 1 and 2].

Close

Figure 1:

Fiberoptic endoscopic evaluation of swallowing (FEES) instruments – entire setup.

Export to PPT

Close

Figure 2:

Fiberoptic endoscopic evaluation of swallowing (FEES) final report showing penetration for liquids with bolus residue.

Export to PPT

Analysis included PAS^[12] and BRS^[13] scoring, providing insights for management recommendations like dietary adjustments.

RESULTS

A total of 41 patients were enrolled: two were lost to follow-up, one patient had Ryles tube insertion done during RT, and one patient deteriorated in the course of RT. So, finally, 37 patients were taken for analysis [Table 1].

We used Cox regression models to predict the time for eating assessment tool, penetration aspiration scale, and bolus residue scale scores to return to mild to normal range after RT or concurrent chemoradiation. We utilized univariate and multivariate Cox proportional hazards models, considering a p-value < 0.05 as statistically significant. Due to a one year study period, follow-up was limited to six months, with event/censored data considered until 120 days for parameters to return to normal.

Results based on normalization of the scores witgh respect to the covariates such as age, gender, performance status,diagnosis, staging, radiotherapy details and baseline scores. Cox regression evaluated age, eating assessment tool at baseline (EATBASE), penetration aspiration scale at baseline (PASBASE), bolus residue scale at baseline (BRS) as both continuous and categorical variables, while other covariates were considered categorical. The time for the scores to return back to normal has been recognized as a surrogate measure of the quality of life, indicating that less time for scores to normal leads to improved swallowing capability and better patient quality of life.

Time taken for eating assessment tool (EAT) score to return to normal as a dependent variable (“EATTIME”)

Here, we have a likelihood ratio of the fit of the full model relative to a null (intercept only) model. Statistical significance suggests that the model is a significant improvement [χ ^2[12] = 37.585, p < 0.001] in fit relative to the null. This indicates that at least one population regression coefficient is different from 0. A log minus log plot for the variables of interest is usually plotted to check whether the proportional hazards assumption holds good and is not violated.

In Figure 3, we can see that the lines are parallel to one another for the variable “Age Categorical”, age as a categorical variable and hence the proportional hazards assumption holds good. This was checked for all variables of interest to make sure that the proportional hazards assumption was not violated.

Close

Figure 3:

Log minus log of EAT TIME – Age Cat. EAT: Eating assessment tool, AGECAT: Age categorical, LML: Log minus log.

Export to PPT

In Table 2, hazard ratio for EATBASE C CATEGORIAL (event vs censored) is 0.079 (95% CI, 0.01–0.52; P = 0.009), signifying a 7.9% longer time for eating assessment tool score normalization if the baseline is moderate to severe (statistically significant). hazard ratio for eating assessment tool at baseline continuous is 0.82 (95% CI, 0.71–0.95; P = 0.01), indicating 17.2 days increase for EATSCORE normalization with a 1-point increase (statistically significant).

Table 2: Eating assessment tool – significance and hazard ratio.

Variables	Hazard ratio (95% CI)	P-Value
Nasopharynx and oral cavity		0.49
Oropharynx and larynx	2.29 (0.40–12.95)	0.34
Hypopharynx	1.07 (0.22–5.10)	0.92
Stage	0.80 (0.27–2.33)	0.68
RT technique	1.09 (0.41–2.93)	0.85
RT dose	3.27 (0.98–10.92)	0.05
Chemotherapy	1.40 (0.47–4.15)	0.54
Age continuous	0.89 (0.81–0.98)	0.02
Age cat (0), <40		0.73
Age cat (1), 41–60	0.22 (0.004–12.43)	0.46
Age cat (2), >60	0.62 (0.08–4.39)	0.63
Gender (Male vs Female)	2.27 (0.81–6.36)	0.11
Eat base (Categorical)	0.07 (0.01–0.52)	0.009
Eat base continuous	0.82 (0.71–0.95)	0.01

RT: Radiotherapy, CI: Confidence interval, Age cat: Age categorical, Eat: Eating assessment tool, Bold value indicates its statistically significant.

Hazard ratio for age continuous is 0.89 (95% Confidence Interval, 0.81–0.98; P = 0.02), implying 10.4 days increase for EAT normalization with a one year age increase (statistically significant). Radiotherapy dose as a categorical variable (0=66 Gray, 1=70 Gray) has HR 3.27 (95% Confidence Interval, 0.98–10.92; P = 0.05), suggesting 3.28 times longer for EAT scores to return to normal with 70 Gy compared to 66 Gy (marginally significant). Other variables like types of cancer, stage, chemotherapy, performance status, and RT technique were not statistically significant.

DISCUSSION

Head and neck cancer is globally recognized as the seventh most prevalent form of cancer. For patients with locally advanced head and neck cancers, radiotherapy or concurrent chemo-radiotherapy has emerged as the established standard of care, offering improved outcomes in terms of local control and overall survival.^[1] Swallowing impairment frequently follows RT for head and neck malignancies, significantly impacting nutrition and overall quality of life (QOL).^[14]

Logemann et al. emphasize that pretreatment dysphagia severity is influenced by tumor stage and site, with significant impacts on specific swallowing disorders.^[15] Liou et al. stress the need for combined objective and subjective dysphagia assessments, highlighting varying outcomes in oropharyngeal and nasopharyngeal tumors. They identify advanced age, multiple head and neck cancer (HNC) diagnoses, and radiotherapy as predictors of unfavorable scores.^[16] Xinou et al.’s study on HNC patients post-chemo-radiotherapy (CRT), using video fluoroscopic swallow study (VFSS) and modified barium swallow impairment profile (MBSImP), underscores persistent severe swallowing deficits and aspiration risks, urging comprehensive evaluations for better understanding.^[17]

Dysphagia and aspiration are recognized as fatal side effects. These factors all pointed to the requirement for an objective analysis of these functions. This was evaluated using the fiberoptic endoscopic evaluation of swallowing (FEES) technique. In our study, the main objective of the study is to assess the swallowing subjectively and objectively.

The eating assessment tool-10, penetration aspiration scale, and bolus residue scale scores were analyzed statistically. All these scores showed a definite improvement from baseline (Before Radiotherapy) to three months post-radiotherapy. Furthermore our study predicts the hazard ratio, from which we can predict the exact percentage and number of days the scores to come back to normal, with one variable compared to the other.

After multivariate statistical analysis, radiation dose and age were of statistical significance as continuous variables in the eating assessment tool-10 score. For every year increase in age, the duration of time taken for EAT-10 to come back to normal increased (approx. 10.4 days). Patients who received 70 Gray took longer time (approx. 3.28 times) to come back to normal than who received 66 Gray.

Similarly, in multivariate statistical analysis for penetration aspiration scale scores, gender and age were of significance as categorical variables. Males took a longer time (approx. 4.6 times) compared to females and patients in the category 41–60 years of age took less number of days (8.9% less) for the scores to return to normal compared to >60 years age.

In the bolus residue scale category, age and gender were of statistical significance. For every year increase in age, it took a longer time for the BRS score to come back to normal (1.12 times). BRS score for males (3.5 times more) took a longer time to revert back to normal. Dysphagia/Aspiration at risk structures were contoured and analyzed in both the techniques volumetric modulated arc therapy (VMAT) and intensity modulated radiotherapy (IMRT). Since it did not show statistical significance, it was not depicted in our thesis. Similarly, the baseline to one month category and one month to three month category analysis was done. Since they did not show any statistical significance, they were not projected in the paper.

CONCLUSION

In our model, it is possible to predict the number of days it takes for the eating assessment tool-10, penetration aspiration scale, and bolus residue scale scores to return back to normal. We can identify these subsets of patients who are prone to develop complications like swallowing difficulties and aspiration so that we can initiate early preventive measures like swallowing exercises, nutritional advises, and Ryles tube insertion at the earliest. By implementing these measures earlier in the treatment process, we can enhance the quality of life for these patients, thereby improving their overall well-being and functional outcomes.