Quantitative Analysis of Temporal Parameters Correlated with Aspiration and Lesion Location in Stroke Patients

Dysphagia, difficulty of swallowing, is common in acute stroke patients. It has been reported to affect from 8.1 to 80% of post-stroke patients [1, 2]. This large variance is due to differences in the definition of dysphagia, the method of assessing swallowing function, and the timing of swallowing assessment after stroke. As dysphagia can lead to serious medical conditions including malnutrition, dehydration, aspiration pneumonia, and choking, it is important to accurately recognize and diagnose dysphagia after stroke [2‐4].

Several diagnostic methods such as bedside swallowing tests, videofluoroscopic swallowing studies (VFSS), and fiberoptic endoscopic evaluation of swallowing (FEES) have been used for detection for dysphagia after stroke. VFSS and FEES both provide structural and functional information of the swallowing. However, VFSS is more useful examining entire swallowing process, because FEES has a limitation in observing pharyngeal phase problem due to the occurrence of a white-out period during some part of the pharyngeal phase. So VFSS has been considered as a gold standard examination [5‐7]. Although it is a widely used assessment, reports of its validity and reliability of measurement are varied [8, 9]. To achieve a more objective and quantitative analysis for VFSS, event time of bolus and structural displacement during swallowing are analyzed for both clinical and research purposes [10‐14]. Many temporal analyses of swallowing events using VFSS have been performed on normal adults [10, 11, 13, 15‐21]. Although there were some studies analyzing temporal parameters in VFSS of stroke patients, most focused on a few specific parameters, and the definition of temporal parameters is often inconsistent from study to study [15, 22‐34]. Therefore, consistent and detailed data on temporal parameters, which are considered clinically valuable, are needed for reliable quantitative analysis of VFSS.

The purpose of this study was to identify differences in temporal parameters depending on the presence of aspiration and the severity of penetration-aspiration scale (PAS) in stroke patients. Also, we investigated whether there was a significant difference in temporal parameters by stroke lesion location.

We demonstrated that prolonged PRT, LVC duration and UES opening duration were negatively correlated with swallowing function with stroke patients indicating significantly increased risk of aspiration. In addition, oral phase duration and UES opening duration were significantly prolonged in supratentorial and infratentorial lesion of stroke, respectively.

Many previous temporal analytical studies focused on pharyngeal delay time (PDT) or pharyngeal transit time (PTT) [17, 25‐31, 38‐42], whereas few studies focused on pharyngeal response time (PRT). The term PRT has been used interchangeably with PDT [41] or defined as a completely different time interval [17, 38‐40]. In our study, we used PDT and PRT separately as defined by Logemann et al. [17, 39, 40]. Previous studies suggested that prolonged PDT is associated with aspiration [18, 26‐28, 42]. Prolonged PTT was also reported to increase the risk of aspiration [25, 29‐31]. However, in our study, there were no significant differences in PDT or PTT between the aspirators and non-aspirators, but PRT was significantly longer in the aspirator group compared with the non-aspirator group (Tables 2, 3). To the best of our knowledge, importance of PRT was first demonstrated in our study. In addition to PDT and PTT, which were suggested as important parameters in the pharyngeal phase in the previous studies, it would be helpful to analyze dysphagia in detail if PRT is included in temporal analysis.

When contrast enters laryngeal vestibule, this is called penetration, which can lead to aspiration that enters beyond the vocal cord. Therefore, it is important that laryngeal vestibule closure (LVC) occurs properly in an appropriate timing in the pharyngeal phase to prevent penetration or aspiration [32, 43‐45]. Two temporal parameters were analyzed for the laryngeal vestibule closure: LVC reaction time (how quickly the laryngeal vestibule get closed at the beginning of the pharyngeal phase) and LVC duration (how long the laryngeal vestibule remains closed). Regarding the LVC reaction time, no significant differences were observed in our study depending on the presence of aspiration. There were some studies in which LVC reaction time was significantly prolonged in aspirators. Kahrilas et al. [44] suggested that prolonged LVC reaction time and UES reaction time were key abnormalities leading to laryngeal penetration or aspiration. Cabib et al. [32] also suggested that prolongation of total oropharyngeal swallow response along with prolonged LVC reaction time and UES opening reaction time were seen in stroke aspirators. Park et al. [23] reported that initiation of laryngeal closure was significantly prolonged in the stroke aspirators compared with stroke non-aspirators or normal subjects. In these previous studies, the starting point of the LVC reaction time was set as BPM or glossopalatal junction opening, whereas in our study, it was set as HYB. However, if the starting point of the LVC reaction time is set to the same point as in previous studies, it is difficult to purely measure the time it takes for the laryngeal vestibule to close because the PDT is included in the LVC reaction time. Therefore, we set the starting point of the LVC reaction time as HYB, based on the study of Steele et al. [10, 16]; our settings have led to different results from the previous studies.

In this study, prolonged LVC duration was significantly associated with the presence of aspiration or PAS severity. Prolonged LVC duration may be caused by prolonged swallowing time. In our results, aspirators showed prolonged pharyngeal phase (PTT), although this was not statistically significant, as well as significantly prolonged LVC duration compared to non-aspirators. However, this result was not consistent with previous studies. Power et al. [27] reported that there was no significant difference in LVC duration between normal subjects, stroke non-aspirators, and stroke aspirators. Park et al. [23] reported that LVC duration was significantly shorter in stroke aspirators compared with normal subjects. The inconsistency of the results of these studies, including ours, indicates that it is difficult to predict aspiration risk simply by the length of the LVC duration. Rather, appropriate timing and length of laryngeal vestibule closure are important parameters to investigate regarding aspiration prevention.

The UES should be opened at the appropriate timing for efficient and safe swallowing. If the UES does not open appropriately, bolus transition from pharynx to esophagus become difficult, which can lead to a large amount of pharyngeal remnant and result in aspiration. Some previous studies have discussed the relationship between UES opening duration and aspiration. Kim et al. [33] found that prolonged UES opening duration is related to aspiration, which results in retrograde flow increasing aspiration risk. They suggested that UES stays open longer due to prolonged bolus transition, slower UES muscle contraction and hyolaryngeal excursion after stroke. Consistently, in our study, prolonged UES opening duration was associated with aspiration and higher PAS score. In contrast, Molfenter and Steele [24] suggested that in stroke patients, aspirators demonstrated significantly shorter UES opening duration compared with non-aspirators. The reason their results are not consistent with ours is likely because they set the definition of their aspiration group to PAS 3 or higher, which is different from the commonly used criteria. Therefore, the results of our study, which set the criteria of aspiration as PAS 6 or higher, can be considered more clinically meaningful results.

Although there were some studies comparing the pattern of dysphagia between supratentorial and infratentorial strokes, few studies quantitatively compared temporal parameters of VFSS. Daniels et al. [46] identified that a higher PAS score is more directly associated with infratentorial stroke than supratentorial stroke, but they did not analyze the temporal parameter. A previous study by Kim et al. [34] reported that the pharyngeal remnant and PAS score were more severe in infratentorial stroke patients than in supratentorial stroke patients. However, they suggested that there were no significant differences in the PTT and functional dysphagia scale. Similarly, in our study, there was no significant difference in the PTT by stroke lesion location.

Previous studies have reported that the oral phase of swallowing is affected by cognitive function [47‐49]. Jo et al. [48] reported that cognitive function had a meaningful influence on the oral phase in supratentorial stroke patients. Moon et al. [47] found that the coordination of the lips, tongue, and oropharynx was associated with the degree of cognitive impairment and stroke severity. Consistent with these results, in our study, we found that oral phase duration was significantly prolonged in supratentorial stroke patients compared with infratentorial stroke patients.

Similarly, in our study, UES opening duration was significantly prolonged in the infratentorial lesion group compared to the supratentorial group. Lower brainstem is important for swallow response in pharyngeal phase as central pattern generator and adjacent medullary reticular formation are situated. Therefore, infratentorial stroke, such as lateral medullary syndrome, can result in UES dysfunction [32, 50‐52]. Steinhagen et al. [50] suggested that brainstem stroke patients may experience UES and pharyngeal incoordination, UES abnormal relaxation, and closure. Another study suggested that up to 80% of patients of brainstem stroke have dysphagia by failed UES relaxation [51]. However, in the present study, patients with severe UES dysfunction were excluded due to difficulty in temporal analysis, so the results may be less accurate. Future studies that include these patients may demonstrate different results.

This study has some limitations. We used fluoroscopy machine of 15 FPS unlike many previous studies that used 30 FPS machine. Although this allowed us to reduce the radiation dose for the patients, the results of temporal analysis could have been less accurate. Since this was a retrospective and cross-sectional study, this may have resulted in imbalance in the number of samples between groups, inconsistency in the time after stroke onset and the distribution of dysphagia severity. In addition, we did not have any control group to compare the results of stroke patients. Therefore, a prospective study with a large number of balanced samples between groups, including the control group, should be performed in the future. Moreover, subjects who could not progress from the pharyngeal phase to the esophageal phase due to severe UES dysfunction were excluded from this study because fully measuring temporal parameters included in pharyngeal and esophageal phases was impossible in these cases. Because many of these patients excluded had infratentorial lesions, a sufficient number of infratentorial lesion group subjects could not be recruited. In addition, only the videos swallowing thin liquid 2 cc were used for analysis because there was no previous barium residue present since in our hospital protocol, thin liquid 2 cc is tested first. Similarly, only the videos of first subswallow were included in this study, because it was easy to accurately analyze the videos in which existing barium residue does not remain in the oropharyngeal area. However, as many post-stroke patients have multiple subswallows per bolus and more severe penetration or aspiration could occur on additional subswallows. Therefore, this may have resulted in underestimation of the severity of penetration or aspiration. If further study including not only various volumes and viscosities of materials but also analyzing multiple subswallows, more accurate and clinically applicable results could be obtained.

This study found that temporal analysis of VFSS can be a clinically valuable tool to help understand the dysphagia patterns related to aspiration and location of brain lesions in stroke patients. More accurate comparative analyses will be possible if subjects are recruited prospectively and a control/non-stroke group is included. Follow-up exams will also need to be performed in future studies.