LONG-TERM RESULTS OF ENDOSCOPIC PYLOROMYOTOMY: FOLLOW-UP OF A RANDOMIZED, SHAM-CONTROLLED TRIAL (GREG)

Aims
We previously reported 6 months results of a prospective randomized sham controlled trial comparing endoscopic pyloromyotomy (G-POEM) with a sham procedure in patients with gastroparesis (GREG). G-POEM was superior to sham in terms of symptoms improvement and gastric emptying rate. Moreover, crossover G-POEM performed after a sham procedure was effective in 75% of patients. All patients have been prospectively followed according to an original protocol. Here, we report the long-term results. Unblinding of patients occurred 6 months after the originally allocated procedure.

Methods
In GREG trial, we included patients with refractory and severe gastroparesis (gastroparesis cardinal symptom index (GCSI) > 2.3, duration of symptoms at least 6 months, with no adequate response to conservative measures). Patients randomized to the sham group with persistent symptoms were offered cross-over G-POEM. Symptoms were assessed in all enrolled patients (including those with a sham procedure only) at 12, 24 and 36 months, gastric emptying study and upper GI endoscopy were repeated at 12 months only in patients who underwent G-POEM (primary or cross-over). Main outcomes were proportion of patients with treatment success and recurrences. Kaplan-Meier curve has been constructed to analyze the long-term effects.

Results
From a total of 41 randomized patients (17 diabetic, 13 postsurgical, 11 idiopathic; 46% male), 33 patients received G-POEM (21 primary, 12 cross-over). Among these, median follow-up was 24 months and treatment success rates evolved from 71% (95% CI: 59%-90%) at 6 months to 65% (95% CI 52%-85%) at 12 months, 62% (49%-82%) at 24 months and 59% (95% CI 44%- 78%) at 36 months. The corresponding numbers for diabetic, postsurgical and idiopathic gastroparesis at 36 months were 71%, 36% and 65%. Overall, 4 patients (2 diabetic, 1 postsurgical, 1 idiopathic) with previous treatment success experienced a recurrence of symptoms, however none of these patients required further intervention yet. One female patient with a primary treatment failure underwent redo G-POEM with treatment success (postsurgical etiology). Seven patients underwent a sham procedure only, 4 achieving treatment success at 6 months with no symptomatic recurrencies up to 36 months follow-up. Among three “only sham” patients without treatment success, one patient died from other reason and two patients who initially refused cross-over G-POEM decided to undergo this procedure later (one with treatment success, one scheduled).

Conclusions
Endoscopic pyloromyotomy leads to a sustained improvement in approximately two thirds of patients with refractory GP and the rate of recurrencies is low. Diabetic and idiopathic GP showed a trend towards a higher long-term effect compared to postsurgical GP. Significant improvement of symptoms after sham procedure was infrequent but long-lasting.