In this issue of Endoscopy, Gonzalez et al. present the first double-blind randomized controlled trial to compare the efficacy of gastric peroral endoscopic myotomy (G-POEM) versus botulinum toxin injection (BTI) for the treatment of refractory gastroparesis [1]. Gastric peroral endoscopic myotomy (G-POEM) was first introduced in 2013 as an alternative option for patients with refractory gastroparesis [2]. Recently, two clinical trials, a large multicenter prospective trial by Vosoughi et al. [3] and a randomized sham-controlled trial by Martinek et al. [4], have evaluated the use of G-POEM in refractory gastroparesis and demonstrated a clinical success rate of 56%–70% at 12 months, with low rates of adverse events. Despite these promising results, G-POEM remains a technically challenging procedure, available only in expert centers, while BTI is widely available. Previously no study had directly compared the two endoscopic pylorus-directed therapies.

In this study, Gonzalez et al. enrolled 40 patients with at least a 1-year history of gastroparesis, refractory to at least 6 months of optimal medical therapy. Clinical success was defined as a >1-point decrease in the total Gastroparesis Cardinal Symptom Index (GCSI) score. Although the authors demonstrated a numerically higher rate of clinical success on intention-to-treat analysis in the G-POEM group compared with the BTI group at 3 months (65% vs. 40%; P=0.10), statistical significance was not met. Medium-term durability was maintained in both groups, with clinical success rates at 12 months in the G-POEM and BTI groups of 60% and 40%, respectively (P=0.20). Gastric emptying at 3 months improved in 72% of the G-POEM group compared with 50% of the BTI group (P=0.17).

The major differences between the current study and the aforementioned studies are worth noting. Gonzales et al. demonstrated a clinical success rate slightly lower than that of Martinek et al., but higher than that of Vosoughi et al. This is likely multifactorial. First, Martinek et al. had the highest proportion of diabetic patients, in whom G-POEM has been shown to be more effective [4] [5]. We now know that the largest etiologic group is in fact idiopathic [6], and therefore the study of Gonzalez et al. may be more reflective of real-world outcomes. Second, the definition of clinical success was different across the studies. Vosoughi et al. defined clinical success as a ≥1-point decrease in the total GCSI, with >25% decrease in at least two of the subscales. Martinek et al. defined treatment success as a decrease of ≥50% in the total GCSI score. The current study modified their definition of clinical success from a decrease in GCSI of >0.6 to a decrease of >1, reflecting the rapidly evolving literature on G-POEM from the time of study conception to publication. These diverse definitions of clinical success contribute to the variable findings in each paper, and clearly highlight a need to standardize objective evaluation of treatment response.

Despite an essentially negative outcome, the study of Gonzalez et al. is not without noteworthy findings. It confirms G-POEM to be at least modestly effective, durable, and safe for the treatment of refractory gastroparesis. The study should be commended for its many strengths, which include its robust double-blind trial design, which is of even greater importance given the functional nature of gastroparesis. Furthermore, the group used the validated GCSI score and quality of life indicators, including the GIQLI and SF-12. Follow-up was vigorous, with four visits in the first 12 months and the use of gastric emptying studies before and at 3 months post-procedure.

However, limitations should also be noted. First, as discussed earlier, the group adjusted the definition of treatment response in line with the evolving literature, but adopting a more rigorous efficacy definition may have resulted in a study underpowered to demonstrate statistical significance. Second, a gastric emptying study was used as an objective marker of treatment response. Gastroparesis is however a heterogeneous disease and the ideal measure of response is unknown. The accepted “gold standard” of diagnosis – i.e. delayed solid gastric emptying – is often bidirectionally labile (i.e. potentially worsening or improving over time) and can frequently occur independently of changes in symptoms [7]. Pyloric distensibility measurements using a functional lumen imaging probe are limited and still evolving in gastroparesis, but are gaining momentum. Both pyloric diameter and the distensibility index have been shown to increase after G-POEM and may be associated with short- to medium-term clinical response [8] [9] [10]. These were not explored in the current paper, presumably because the use of distensibility measurements at the pylorus was still in its infancy at the time of study conception, but this would be a useful addition to any future trials. Finally, despite their rigorous follow-up schedule, there was loss during follow-up, particularly in the BTI group. Nevertheless, as the authors correctly pointed out, the intention-to-treat analysis should offset this issue.

To conclude, this relatively small randomized controlled trial strengthens the position of pylorus-directed therapies in the treatment of refractory gastroparesis. While G-POEM was not shown to be superior to BTI, it remains an effective and safe option in expert centers. Further blinded comparative trials with objective parameters of treatment response, such as distensibility measurement, should be encouraged.

Article published online:
15 March 2024

© 2024. Thieme. All rights reserved.

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany