Abstract 17857: Trial Duration And Relative Risk Of Clinical Worsening With Combination Therapy In Pah Patients: A Meta-analysis
Introduction: Most randomized controlled trials (RCT) report results as relative risk (RR). The number needed to treat (NTT) has been used alternatively to provide information on the risk reduction in absolute terms. Both RR and NNT are time-dependant mesures.
Hypothesis: A systematic review and meta-analysis was performed to assess if trial duration influenced the RR and absolute risk (RA) of worsening in RCT comparing combination (CT) of PAH-specific therapies versus monotherapy (MT).
Methods: We searched MEDLINE, EMBASE and the Cochrane Library from January 1990 to May 2015 for RCT comparing CT to MT in PAH. The primary outcome was the risk of clinical worsening. RR and NNT were estimated at 16, 26, 52 and 104 weeks using the Kaplan-Meier method using data from recent long-term event-driven trials. We assessed if trial duration correlated with RR and NNT using weighted meta-regression with mixed effects for NNT.
Results: 4038 patients (15 studies) were included. CT has shown a risk reduction for clinical worsening (CW) (risk ratio [RR]:0.65;95%CI0.58-0.73; P<0.00001), with treatment effect consistent across subgroups. The RR correlated with trial duration (β=0.0035,R2=1.00,p=0.011), but the NNT did not, being relatively constant (mean NNT≈8,β=-0.15,p=0.13). In long-term event-driven trials, the mean NNT decreased until 52 weeks of follow-up, being stable thereafter (mean NNT≈7). The mean RR increased from 0.40 at week 16 to 0.42, 0.49 and 0.62 at weeks 26, 52 and 104.
Conclusions: CT has shown a significant reduction in CW compared to MT in PAH. Individual RR positively correlated with trial duration. In long-term event-driven trials, the RR initially increased and the NNT decreased with increasing follow-up duration and were constant after 12 months, implying a relatively constant absolute risk reduction thereafter. This constant absolute risk reduction after 6-12 months of follow-up questions the requirement for long-term event-driven trials in PAH.
Author Disclosures: A. Lajoie: None. G. Lauziere: None. J. Lega: None. Y. Lacasse: None. S. Martin: None. S. Simard: None. S. Bonnet: None. S. Provencher: None.
- © 2016 by American Heart Association, Inc.