Dysphagia Outcomes in Zenker Diverticulum: A Longitudinal POuCH Study

ABSTRACT

Objective

Patients with cricopharyngeus muscle dysfunction (CPMD) with and without diverticula (e.g., Zenker Diverticulum) often struggle with dysphagia for years prior to diagnosis or intervention. Surgical treatment is successful; yet, there is limited long‐term data on dysphagia outcomes after surgery.

Methods

Individuals prospectively enrolled in the Prospective Outcomes of Cricopharyngeal Hypertonicity (POuCH) collaborative with and without diverticula who underwent surgery from November 2014 to August 2024 and had at least 12 months of follow‐up were included. Eating Assessment Tool 10 (EAT10) was used to characterize patient‐reported outcomes. Descriptive statistics were performed using means, frequencies, and spaghetti plots.

Results

Of 164 patients initially identified, 2 withdrew and 2 were excluded for incomplete data; 160 were included. 31% were women, with a mean (SD) age of 70.1 (11.5). Endoscopic surgery (113, 71%) was more common than open technique (47, 29%). 19.4% of patients had undergone previous surgery. Postoperatively, patients were found to have improved EAT‐10 scores, with smaller changes recorded after 36 months. Median change in EAT‐10 compared to preoperative score was −12 points [IQR: −18, −5] at 12 months. 39 (24%) patients were followed for > 24 months (median change −11 [IQR: −14, −7]); 26 (16%) > 36 months (−9.5 [−17, −5]), and 12 (7.5%) > 48 months (−4 [IQR: −18, 4]). Six patients died during the study period.

Conclusion

Improved dysphagia after surgery appears stable with minimal change over time. Most patients (96.3%) survive at least 1 year. Surgical intervention for patients with CPMD with or without diverticula is an effective and reliable treatment option.

Level of Evidence

4.

Surgery for cricopharyngeal muscle dysfunction, including Zenker's diverticula, has been shown to improve quality of life immediately post‐operatively. We present the first analysis of long‐term outcomes using validated, patient‐reported outcome measures.

1. Introduction

Zenker diverticulum (ZD) and cricopharyngeus muscle dysfunction (CPMD) are disorders of cricopharyngeus relaxation, increasing hypopharyngeal pressure and causing (in cases of ZD) herniation of a pseudo‐diverticulum [1, 2, 3, 4]. Most patients are offered surgery; endoscopic approaches are often chosen over open treatment [5, 6, 7, 8, 9]. Much of the population that undergoes surgery for ZD and CPMD can be older or even have malnutrition secondary to an acquired dysphagia. However, prior data suggest that surgical intervention is relatively safe, with great opportunity for symptomatic improvement [10, 11, 12, 13].

Endoscopic surgical techniques have been common since the initial publication from Van Overbeek in 1980 [14]. More recent work shows that most surgeons now prefer transoral endoscopic techniques, either performed with laser, electrocautery, or stapler [2, 6, 8, 9, 15]. In addition to rigid endoscopic techniques, peroral endoscopic myotomy (POEM) is a flexible endoscopic approach to the management of Zenker diverticula with emerging supportive data in the gastroenterology literature [16, 17, 18]. In 2002, a charge analysis of endoscopic staple diverticulostomy and transcervical diverticulectomy showed roughly equivalent charges with a decreased time in the operating room and a shorter inpatient stay in the endoscopic cohort [19]. Both endoscopic and open surgical management of ZD provide significant immediate improvement in dysphagia symptoms, with 66% patients reporting a 100% improvement and 88% with greater than 50% improvement regardless of transoral or transcervical approaches [20].

The Prospective Outcomes of Cricopharyngeus Hypertonicity (POuCH) collaborative is a multi‐institutional (11 sites) group that, since 2017, has aimed to collect and study adult (≥ 18 year old) patients with cricopharyngeus muscle dysfunction (CPMD) with and without hypopharyngeal diverticulum [6, 7]. Notably, the POuCH collaborative uses validated patient‐reported outcomes, fluoroscopy, standardized demographic, surgical, and complications within a multi‐institutional RedCAP database with a broad geographic range within the United States and including New Zealand [6, 7, 21, 22]. This is a fundamental difference from the majority of prior studies, which have used nebulous definitions of dysphagia to describe longitudinal outcomes retrospectively. Previous work on longitudinal outcomes of repair is limited to single institutions with relatively small cohorts [23, 24]. The purpose of this study is to examine the longitudinal effect of surgery on patients with ZD and CPMD across multiple institutions with validated patient‐reported outcome measures, specifically the Eating Assessment Tool‐10 (EAT‐10) [21]. We hypothesize that patients would have decreased improvement in EAT‐10 due to other comorbid conditions contributing to dysphagia, such as esophageal dysmotility or hiatal hernia [25].

2. Methods

This is a prospectively collected multi‐institutional database of patients with cricopharyngeus muscle dysfunction with and without diverticula. Approval of the study (IRB 2016–6955) was obtained from the Institutional Review Board at all participating institutions. All participants are 18 years of age or older and have been individually enrolled and consented by their home institution for inclusion into the POuCH database. In this study, we included patients with CPMD with or without diverticula and undergoing surgery with at least 12 months follow‐up. Eating Assessment Tool 10 (EAT10) was used to characterize patient‐reported outcomes [21]. Demographics including age, sex, time of diagnosis, time of surgery, and time to follow‐up were included. Parametric data were summarized with means and standard deviations; non‐parametric information (most notably EAT‐10) were summarized with median and inter‐quartile range.

Patients were followed via the surgeon's standard practice for the first 12 months, either via in‐person or telehealth visits. After this, if discharged from the practice of the operating surgeon, patients were contacted via telephone and email from the collaborative to provide further follow‐up information. We excluded patients with incomplete data (e.g., missing EAT‐10 post surgery). Descriptive statistics were performed using parametric and nonparametric measures as appropriate, as well as spaghetti plots for visualization.

3. Results

3.1. Patient Demographics and Follow Up

A total of 160 patients were included. Four additional candidates with 12 months of data were identified; two had withdrawn from the study, and two did not have complete EAT‐10 data post‐operatively. In the remaining cohort, there were 49 women (30.6%) and 111 men (69.4%). The mean age was of 70.1 (standard deviation [SD] 11.5). Thirty‐one patients had undergone previous surgery (19.4%). Symptoms have been long‐standing in many cases, with only 25 patients (15.6%) having symptoms for less than 1 year (Table 1). However, nearly a quarter of patients described unintentional weight loss in the year prior to surgery (n = 36, 22.6%), with mean reported weight loss of 15.7 lbs (SD 10.8).

TABLE 1.

Demographic data and characteristics of the study population.

Characteristics (N = 160)	Value
Age at surgery, years, mean (SD)	71.0 (11.7)
Gender, number (%)
Female	49 (30.6)
Male	111 (69.4)
Pre‐operative EAT‐10, median [IQR] a	15.0 [9.0, 25.0]
Any neuromuscular disorder, number (%)	7 (4.4)
History of Zenker surgery, number (%)	31 (19.4)
Any unintentional weight loss over the past year? Number (%)	36 (22.6)
Reported weight loss, pounds, mean (SD) a	15.7 (10.8)
Duration of symptoms, number (%)
< 1 year	25 (15.6)
1–3 years	68 (42.5)
3–10 years	38 (23.8)
> 10 years	29 (18.1)
Size, number (%)
< 1 cm (CP bar)	17 (10.8)
1–2 cm	74 (47.1)
2–3 cm	5 (3.18)
3–4 cm	49 (31.2)
5–6 cm	9 (5.7)
Other (6.8, ″large)	3 (1.91)
Unknown	3
Type of Surgery (index surgery), number (%)
Endoscopic	114 (71.2)
Open	46 (28.8)
Intraoperative diagnosis, number (%)
Isolated CPMD	19 (12.0)
Zenker	125 (79.1)
Killian Jameson	9 (5.7)
Laimer	2 (1.3)
Other (traction, iatrogenic)	3 (1.9)
Unknown	2

^a EAT‐10 n = 157, weight loss pounds reported in n = 31.

Median preoperative EAT‐10 score was 15.0 (interquartile range [IQR] 9–25). Three patients did not have a pre‐operative EAT‐10 score to include in the calculated median. A variety of sizes of diverticula and CPMD were represented in the study (Table 1). The plurality were 1–2 cm (n = 74, 47.1) with the next most common group being 3‐4 cm in size (n = 49, 31.2%). Endoscopic surgery (113, 71%) was more common than open technique (47, 29%). Intraoperative findings confirmed a diagnosis of ZD in most cases (n = 125, 79.1%), but there were nine Killian‐Jameson diverticula (5.7%), two Laimer diverticula (1.3%), and three traction or iatrogenic diverticula noted (1.9%). Full characteristics of the study population are found in Table 1.

3.2. Post‐Operative Outcomes

Patients had a mean of 2.94 post‐operative EAT‐10 scores recorded (SD 1.45). All 160 patients had at least one follow up, which was at a mean of 3.54 months (SD 2.9). Post‐operatively, patients were found to have sustained improvement in longitudinal EAT‐10 scores years after surgical intervention of any type (Figure 1). For patients followed for greater than 12 months with adequate data (n = 159), the median decrease in EAT‐10 a year post‐operatively was 12 points (IQR −18 to −5). The decrease was sustained at 24 months (n = 29, −11, IQR −17 to −7) and 36 months (n = 45, −9.5, IQR −17 to −5). At greater than 48 months post‐operatively, there were only 20 patients included, and the median decrease was −4 (IQR −18 to +4) found in Table 2.

FIGURE 1.

Change in EAT‐10 Score over time. Spaghetti plot showing change in EAT‐10 score over time. Each set of lines and symbols demonstrates the outcomes over time of an individual. [Color figure can be viewed in the online issue, which is available at www.laryngoscope.com]

TABLE 2.

Median change in EAT‐10 Score over time.

Follow‐up months	Number	Median [IQR] change in EAT‐10 Score from baseline pre‐operative score
12 months	159	−12 [−18, −5]
> 24 months	29	−11 [−17, −7]
> 36 months	45	−9.5 [−17, −5]
> 48 months	20	−4 [−18, 4]

Six patients died during the study period. No deaths were related to cricopharyngeal muscle dysfunction or surgery. Four patients underwent additional endoscopic surgery during the study period, and three patients underwent subsequent open surgery during the study.

4. Discussion

In this study, we present the first prospective, multi‐institutional study of longitudinal outcomes after ZD and CPMD surgery using validated patient reported outcome measures. Our study confirms that patients do tend to do well post‐operatively and sustain their improvement over time. Only at > 48 months, where the cohort was significantly smaller, was there no difference between post‐operative and pre‐operative outcomes. In the larger cohorts available at 1, 2, and 3 years, there was significant improvement in EAT‐10 scores compared to preoperative measurements.

While surgery is usually initially successful, limited studies have investigated the long‐term success of intervention. In Italy, Bonavina et al. have investigated this twice (2007 and 2015) with the proportion of “asymptomatic” patients measured in 2007 and a Likert scale for severity of dysphagia in 2015. In 2015, 76% of patients over 63 months had improvement based on that outcome measure [26, 27]. Similar rates were described for laser and staple‐assisted surgery in 2007 [27]. A study from the Netherlands reached out to patients via survey a median of 19.2 months after surgery. They found that 92.3% of patients had no further symptoms or improvement from their preoperative baseline [28]. In the United States, Greene et al. assessed long‐term outcomes by telephone interview with symptomatic improvement in 77 patients; 55 patients reported complete resolution at 4 months. Thirty‐nine of 58 surviving patients were assessed long term. At a median of 59 months, patients were often improved, with 22% of patients reporting residual dysphagia, 11% with regurgitation, and 8% with aspiration [29].

Our data affirms the findings of this prior research, but importantly bolsters the strength of the conclusions with its multi‐institutional nature and patient‐reported outcomes intentionally assessed longitudinally for long‐term follow‐up. As noted above, previous work has analyzed patient‐reported outcomes, but limited reports focus on a standardized, validated fashion. EAT‐10 score has become a widely used measure in dysphagia research and is well‐studied for its clinical validity [21]. This study is also unique in its prospective design. Rather than relying on surgeon‐specific practice patterns, enrolled patients also have study‐specific follow‐up that improves the length of follow‐up and power of the findings. Additionally, given its multi‐institutional nature, the study shows that multiple surgeons at multiple institutions achieve sustainable results.

This study does have notable limitations. With regard to its demographics, the study has a significant preponderance of men, which cannot be easily explained and may bias results if women are less likely to have sustained improvement. However, no previous retrospective data has suggested this. Other demographics, including age, are in line with prior studies from other groups [26, 27, 29]. Additionally, there are a relatively small number of cases of isolated CPMD (n = 19, 12%) which may limit exchangeability from the hypopharyngeal diverticula group to the CPMD group.

Further work is needed to better understand the few patients who do not respond to surgery, and those patients who fail to respond over time—that is, the seven patients who required revision. For patients within 1 year of surgery, previous research from the POuCH collaborative has shown that the only preoperative predictor of poor response (< 50% improvement in EAT‐10 score) is the presence of a hiatal hernia [25]. However, further investigation with the longitudinal data will be needed to confirm if this is true over a longer study period. An additional factor that must be considered, given the age of the population in this study, is frailty. Work in large databases has suggested that one proxy measure for frailty, serum albumin, is not associated with outcomes in ZD surgery [11]. Weight loss in the prior year, as measured here, may be a stronger measure to predict outcomes and complications, and requires more research to investigate.

5. Conclusions

In this prospective longitudinal study of patients undergoing surgery for ZD and CPMD, we demonstrate stable outcomes over time using a validated patient‐reported outcome measure. Despite a high proportion of older patients and serious potential complications, mortality after surgery is low, with only six reported mortalities in the cohort across long‐term follow‐up. While prior studies have reviewed long‐term success rates, to our knowledge, no authors have used validated patient‐reported outcomes over a study period of more than a year or in a multi‐institutional fashion. Further work is needed to better investigate individual responses to surgery and determine if any specific groups fail to respond.

Disclosure

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Schuman A. D., McKeon M., Allen J., et al., “Dysphagia Outcomes in Zenker Diverticulum: A Longitudinal POuCH Study,” The Laryngoscope 136, no. 2 (2026): 683–687, 10.1002/lary.70052.