Fluoxetine Shows No Benefit for Post-Stroke Functional Outcomes

December 7, 2018

Despite smaller trials suggesting some benefit, a trial of more than 3000 patients has demonstrated no differences between placebo and fluoxetine 6 months post-stroke.

Martin Dennis, MD

Although the results of several small trials have suggested that fluoxetine might improve patients’ functional outcomes post-stroke, new data indicates that this may not be the case.

The FOCUS trial findings have shown that 20-mg fluoxetine administered daily for 6 months post-acute stroke did not improve functional outcomes (P = .439). It did reduce the incidence of depression after a stroke, but it increased the frequency of bone fractures.1

Study authors, led by Martin Dennis, MD, the chair of stroke medicine at the University of Edinburgh, wrote that “these results do not support the routine use of fluoxetine either for the prevention of post-stroke depression or to promote recovery of function.”

"I think the key take-home message for clinicians is that routine use of fluoxetine to boost recovery after stroke is not effective, and should not be used unless further, more positive evidence becomes available in the future," Dennis and co-chief investigator Gillian Mead, MD, the chair of stroke and elderly care medicine at the University of Edinburgh, told NeurologyLive.

The pragmatic, parallel-group, double-blind, placebo-controlled trial randomly assigned 3127 patients at 103 centers to either fluoxetine (n = 1564) or placebo (n = 1563) and assessed them for functional outcomes, measured with the modified Rankin Scale (mRS), at 6 months. At the end of the study, data were available for 1553 patients in each group (99.3%).

Distribution across the mRS categories at 6 months was similar for both groups when adjusted for minimization (common odds ratio [OR], 0.951; 95% CI, 0.839 to 1.079; P = .439). Additionally, the unadjusted findings provided similar results (common OR, 0.961; 95% CI, 0.848 to 1.089; P = .531). Dennis and colleagues wrote that “the ordinal analysis was done with the assumption of proportional odds, in the model of mRS by treatment. This assumption was found to hold in the score test for proportional odds assumption (P = .9947).”

Of the variables used in the minimization of data, no significant differences between subgroup were observed. The probability of being alive and independent at 6 months (P = .3259), delay from onset to randomization (P = .9507), presence of motor deficits (P = .1530), and presence of aphasia (P = .1234) were all statistically comparable.

"We need more data from the ongoing Affinity and Effects trials to allow us to properly balance the benefits on enhanced mood against the risks in terms of adverse effects," Dennis and Mead said. "These trials should report their results in 2020. It would be good if we could also understand the discrepancy between the results of our FOCUS trial and the myriad small studies which lead us to believe that fluoxetine would be effective."

Secondary end points included scores on the Stroke Impact Scale (SIS) which included multiple functional measurements (including strength, hand ability, mobility, motor functionality, daily activity, physical function, memory, etc.), as well as European Quality of Life-5 Dimensions-5 Levels (EQ5D-5L) score, Mental Health Inventory (MHI-5) score, and vitality from the Medical Outcomes 36-Item Short Form Survey (SF-36) score.

The difference in MHI-5 scores was 76 (95% CI, 60 to 88) with fluoxetine compared to 72 (95% CI, 56 to 88) with placebo (P = .01), though this difference was not sustained through 12 months. Comparatively, both vitality scores (P = .6726) and EQ5D-5L scores (P = .5866) were not significant statistically.

Dennis and colleagues acknowledged that in their assessment, they observed an effect of treatment among the subgroup of patients with motor deficit at baseline (n = 2702) but found no evidence of an effect on the mRS (common OR, 0.919; 95% CI, 0.803 to 1.051; P = .2172) or on motor score based on the mean of SIS Strength, Hand, and Mobility domains (fluoxetine median, 48.43 [interquartile range [IQR], 24.98 to 78.84]; placebo, 52.66 [IQR, 25.28 to 77.22]; P = .4714).

Data did show that patients who were allocated to fluoxetine (n = 210; 13.43%) were less likely than those allocated to placebo (n = 269; 17.21%) to develop new depression by 6 months (difference, 3.78%; 95% CI, 1.26 to 6.30; P = .0033), but these patients also had more bone fractures (45 [2.88%] vs. 23 [1.47%]; difference, 1.41%; 95% CI, 0.38 to 2.43; P = .0070). There were no significant differences in any other event at 6 or 12 months.

"We were not expecting the increased risk of bone fractures associated with fluoxetine treatment. Although some observational studies have suggested this might be an issue, previous [randomized controlled trials] of [selective serotonin reuptake inhibitors] had not confirmed it," Dennis and Mead said. "This is probably because previous [randomized controlled trials] of fluoxetine have been small, with short treatment periods, and predominantly carried out in much younger patients who have much lower risks of falls and fractures. We will be carrying out further analyses to learn more about the fracture risk."

Dennis and colleagues wrote that ongoing trials might be able to confirm the external generalizability of these findings to other patient populations, "and a planned individual patient data meta-analysis could clarify whether any subgroups might benefit from fluoxetine and provide more precise estimates of any harms."

"For now [the results] put the emphasis firmly back on primary prevention—for instance, by ensuring patients with atrial fibrillation are anticoagulated—patients receive thrombolysis and thrombectomy as early as possible and that adequate rehabilitation services are available in both hospital and community settings," Dennis and Mead said.

REFERENCE

1. Dennis M, Mead G, Forbes J, et al. Effects of fluoxetine on functional outcomes after acute stroke (FOCUS): a pragmatic, double-blind, randomised, controlled trial. Lancet. Epub December 5, 2018. thelancet.com/journals/lancet/article/PIIS0140-6736(18)32823-X/fulltext