7CEP THE SAFETY AND EFFICACY OF RECOMBINANT ACTIVATED FACTOR VIIA FOR NON-TRAUMATIC INTRACRANIAL HEMORRHAGE: PATTERNS OF USE AND COMPARATIVE EFFECTIVENESS REVIEW

Tuesday, October 20, 2009
Grand Ballroom, Salons 1 & 2 (Renaissance Hollywood Hotel)
Veronica Yank, MD1, Craig V. Tuohy, BS1, Aaron Logan, MD, PhD2, Dena M. Bravata, MD, MS1, Kristan Staudenmayer, MD1, Robin Eisenhut, BA1, Vandana Sundaram, MPH3, Donal McMahon, MS, MSc1, Ingram Olkin, PhD4, Kathryn M. McDonald, MM1, Douglas K. Owens, MD, MS5 and Randall S. Stafford, MD, PhD2, (1)Stanford University, Stanford, CA, (2)Stanford University School of Medicine, Stanford, CA, (3)VA Palo Alto Health Care System, Palo Alto, CA, (4)Stanford University, Palo Alto, CA, (5)Veterans Affairs Palo Alto Health Care System and Stanford University, Stanford, CA

Purpose: To characterize patterns of off-label rFVIIa use in ICH and evaluate its comparative safety and effectiveness versus usual care.

Methods: Patterns of Use: We analyzed rFVIIa use for ICH using the 2000-08 Premier Prospectives administrative database, a representative sample of 615 US hospitals. Systematic Review: We sought English-language articles comparing rFVIIa versus usual care in ICH from 10 databases (e.g., PubMed, DARE, EMBASE, BIOSIS) through 2/2009. We pooled data from the RCTs for selected outcomes and calculated summary mean differences between rFVIIa and usual care using random effects models.

Results: Patterns of Use: rVIIa use in ICH reached sizable scale in 2004 (250 cases) and increased 8-fold (2,010 cases) by 2008, accounting for 11% of rFVIIa use. Systematic Review: We identified 4,967 citations; 8 studies with 1,023 rFVIIa-treated patients and 551 usual care patients were included (4 RCTs, 3 comparative observational studies, 1 non-comparative observational study). Doses ranged widely: 5-160 mcg/kg. RCTs excluded patients with baseline CT scan >3 hours after symptom onset, oral anticoagulation, deep coma, and anticipated evacuation of the hematoma. Mortality was lower among rFVIIa patients in the RCTs (90-day mortality, 19%) compared to the Premier data (In-hospital mortality, 34%), suggesting the RCT data may be most applicable to less acutely ill patients. Comparing rFVIIa and usual care patients, there was no difference in mortality (arcsine mean difference, -0.051; 95% CI -0.156 to 0.054) or poor functional status (modified Rankin Scale, arcsine mean difference -0.048; 95% CI -0.183 to 0.087). There was a sizable, but non-significant increase in arterial thromboembolism risk with rFVIIa (arcsine mean difference 0.100; 95% CI -0.072 to 0.272). Relative expansion of ICH volume was significantly reduced with rFVIIa compared to usual care (standardized mean difference -0.229; 95% CI -0.384 to -0.074). Hematoma expansion may be reduced with earlier rFVIIa administration, but small sample sizes and study heterogeneity limited our ability to evaluate timing, as well as dose effects.

Conclusions: ICH is a common indication for rVIIa use, but favorable surrogate outcomes may not translate into direct patient outcomes. Until further evidence is available to identify treatment strategies or subpopulations with unambiguous benefit, rFVIIa use in ICH should be approached with caution.

Candidate for the Lee B. Lusted Student Prize Competition