What can we learn from a statistically inconclusive trial? Consensus conference on the EVOLVE study results

The link between serum parathyroid hormone (iPTH) and cardiovascular (CVS) mortality has not been fully elucidated. The EVOLVE Study was designed to test whether a drug such as cinacalcet, aimed at lowering iPTH, could reduce the astonishingly high cardiovascular risk in patients on maintenance dialysis (CKD-5D). Accordingly, the primary outcome of the study was the combined endpoint of time to death or hospitalization due to CVS factors or from any cause. Time to bone fracture and parathyroidectomy were regarded as secondary endpoints. At study completion, the Intention-To-Treat analysis documented a non- significant 7% (Hazard Ratio: 0.93; 95% Confidence interval: 0.85-1.02; P = 0.11) reduction of the primary composite endpoint. However, the intention to treat analysis does not take into account adherence to drug regimens or control for factors that may potentially jeopardize the conduction of the study. In particular, in spite of a careful pre-planned study sample calculation, the final power of the EVOLVE study was 54% instead of the assumed 90%, greatly reducing the reliability of study results. Furthermore, the pre-planned multivariable adjustment of the primary endpoint suggests a nominally significant reduction of the risk of the primary composite endpoint when age is entered into the statistical model. The sensitivity analysis further corroborates this result. The Lag Time Censoring Analysis (LTCA) evidenced a nominally significant 15% risk reduction of the composite endpoint among patients allocated to cinacalcet if the patients follow-up was terminated 6 months after the study drug discontinuation, as pre-planned in the protocol. It is interesting that the LTCA suggests that the effect of cinacalcet weakened over time and became insignificant after about 1 year from drug discontinuation. Although authors could not detect any effect of cinacalcet on bone fracture associated with cinacalcet use, the secondary analyses of the EVOLVE trial suggest a nominally significant 60-70% risk reduction of parathyroidectomy and a reassuring safety profile of prolonged exposure to cinacalcet. In summary, the EVOLVE study adds to the list of inconclusive randomized clinical trials in Nephrology. However, the preplanned exploratory and sensitivity analyses suggest that when imbalances of patients characteristics at study entry (i.e. age) or study drug discontinuation are considered, a 'nominally' significant risk reduction in CVS and parathyroidectomy associated with cinacalcet treatment is noted.