Impact of ARUBA trial on trends and outcomes in symptomatic non-ruptured brain AVMs: A national sample analysis.

[INTRODUCTION] The real-world evolution of management and outcomes of patients with unruptured brain arteriovenous malformations (AVMs) has not been well-delineated following the ARUBA trial findings of no general advantage of initial interventional (surgical/endovascular/radiotherapy) vs. initial conservative medical therapy.

[METHODS] We analyzed the National Inpatient Sample from 2009-2018, capturing 20% of all admissions in the U.S. Validated ICD-9 and -10 codes defined brain AVMs, comorbidities, and the use of interventional modalities. Analyses were performed by year and for the dichotomized periods of pre-ARUBA (2009-2013) vs. post-ARUBA (2014-2018).

[RESULTS] Among the national projected 88,037 AVM admissions, 72,812 (82.7%) were unruptured AVMs and 15,225 (17.3%) were ruptured AVMs. Among uAVMs, 51.4% admitted pre-ARUBA and 48.6% in post-ARUBA period. The post-ARUBA patients were mildly older (median age 53.3 vs. 51.8 (p = 0.001) and had more comorbidities including hypertension, diabetes, obesity, renal impairment, and smoking. Before the first platform report of ARUBA (2009-2012), rates of use of interventional treatments during uAVM admissions trended up from 31.8% to 35.4%. Thereafter, they declined significantly to 26.4% in 2018 (p = 0.02). The decline was driven by a reduction in the frequency of endovascular treatment from 18.8% to 13.9% and inpatient stereotactic radiosurgery from 0.5% to 0.1%. No change occurred in the frequency of microsurgery or combined endovascular and surgical approaches. Adjusted multivariable model of uAVMs showed increased odds of discharge to a long-term inpatient facility or in-hospital death [OR 1.14 (1.02-1.28), p = 0.020] in post-ARUBA. A significantly increased proportion of ruptured AVMs from 17.0% to 23.3% was observed consistently in post-ARUBA.

[CONCLUSION] Nationwide practice in the management of unruptured AVMs changed substantially with the publication of the ARUBA trial in a durable and increasing manner. Fewer admissions with the interventional treatment of unruptured AVMs occurred, and a corresponding increase in admission for ruptured AVMs transpired, as expected with a strategy of watchful waiting and treatment only after an index bleeding event. Further studies are needed to determine whether these trends can be considered to be ARUBA trial effect or are merely coincidental.