Objective : Embolotherapy of progressive or symptomatic neuroendocrine tumor (NET) liver metastases is standard of care in international guidelines, without recommendation regarding embolization technique. RETNET (NCT02724540) is a prospective randomized controlled trial comparing bland, cTACE, and DEB-TACE. Interim safety analyses are planned following accrual of 10 and 30 subjects in each arm. The first safety review is reported here.
Methods : SIR QI guidelines for embolization and TACE set the performance threshold for major complications at 8%, and for any complications 15%. For RETNET, a 20% incidence of serious adverse events as defined by the SIR (D = unplanned increase in level of care; E = permanent adverse sequelae; F = death) was set as the stopping rule for closing an arm. This provides a probability of halting of 62% at the first safety review if the true event rate is 20%. Blinded review was performed by an independent DSMB consisting of three senior interventional oncologists at non-participating institutions after a minimum of 10 subjects in each arm had one-month follow-up after their first embolization cycle.
Results : Two of 11 subjects in the bland arm had SIR Category D events (pain; unrelated pancreatitis 6 months post-embo). The subject with pain required re-hospitalization and withdrew from the study. Two of 10 subjects in the cTACE arm had SIR Category D events (unrelated renal infection, hyponatremia), none resulted in study discontinuation. Four of 10 subjects in the DEB arm had Category D (hyponatremia, carcinoid crisis, both requiring ICU admission) or E (hepatobiliary necrosis; biloma) events, two of which resulted in study discontinuation.
Conclusions : Four of 10 NET subjects embolized with DEB had serious adverse events, two permanent and two resulting in study discontinuation. Based on the trial stopping rule, the DSMB closed the DEB arm.