Medical management of neurogenic lower urinary tract dysfunction (NLUTD) continues to be an important part of guideline-based care for patients with spinal cord injury (SCI) and multiple sclerosis (MS). Anticholinergic based overactive bladder (OAB) medications are the most frequently studied and used medications in this patient population. However, the advent of beta-3 agonists for OAB introduced a new class of medication that could potentially be used to treat patients with NLUTD. An advantage of this class of medication is that their mechanism of action potentially reduces the risk of anticholinergic-mediated side effects. While authors have published systematic reviews and meta-analyses of clinical trial results, there have been no individual patient data meta-analyses (IPDMA); advantages of combining participant level data are increased statistical power, standardized analysis, improved data quality, and avoidance of ecological bias in subgroup analysis, and they are considered the gold standard of meta-analysis. Our objective was to conduct an IPDMA using data from the two published randomized placebo controlled trials(1,2) of mirabegron in people with MS or SCI, and to conduct exploratory subgroup analyses within this patient population. Our hypothesis was that mirabegron would result in significant improvements in the measured outcomes.

Both studies were randomised, multi-center, placebo controlled, trials of mirabegron in people with SCI or MS. The first randomised 32 patients with urge/unaware incontinence on voiding diary to either mirabegron (25mg with a dose titration to 50mg) or placebo for 8 weeks, with a primary outcome of maximum cystometric capacity. The second randomised 66 patients with proven neurogenic detrusor overactivity (NDO) to either mirabegron (50mg) or placebo for 4 weeks after a 2 week placebo wash-in period, with a primary outcome of maximum cystometric capacity.

There were two primary outcomes for this study: the change in maximum cystometric capacity and the change in the patient perception of bladder condition (PPBC). Secondary outcomes were the change in the maximum detrusor pressure, peak pressure of NDO, urodynamic compliance, 24hr pad weight, and an incontinence related quality of life measure (I-QoL).

We conducted three exploratory analyses to test hypotheses based on our clinical experiences with mirabegron in NLUTD. First, we evaluated whether mirabegron had significantly better efficacy among people with SCI or MS; our hypothesis was that mirabegron would have significantly better outcomes among people with MS. Second, we evaluated whether mirabegron had significantly better efficacy among people with SCI ASIA A (complete SCI) versus SCI ASIA B, C, D or E (incomplete SCI); our hypothesis was that mirabegron would have significantly better outcomes among people with incomplete SCI. Third, we evaluated whether mirabegron had significantly better efficacy among people with a high SCI (T6 or higher) versus low SCI (T7 or lower); our hypothesis was that mirabegron would have significantly better outcomes among people with low SCI. 

Statistical analysis

The primary analysis was a comparison between the outcome at the end of study of participants randomised to placebo versus those randomised to mirabegron. Analysis of covariance (ANCOVA) was used to adjust for baseline values. A two-sided confirmatory p-value <0.05 was considered significant for the primary comparison. For our exploratory analyses, we used ANCOVA with an interaction term between randomisation status and the subgroup of interest to formally test for a statistically significant differences between subgroups based on randomisation and an exploratory p<0.25 was considered of interest for future study. Marginal means and 95% confidence intervals (CI) are reported.

In our primary analysis, at the end of study there was a significantly greater improvement in all primary and secondary outcomes aside from urodynamic compliance (table 1). When normal compliance was defined as a binary variable (>20cmH2O), the proportion of the placebo group (77%) and the mirabegron group (88%) that was normal was also similar (p=0.20).

Exploratory analysis of efficacy of mirabegron in SCI versus MS patients
There were 68 patients with SCI, and 30 with MS. There were no statistical interactions among the primary or secondary outcomes that met our threshold of an exploratory p<0.25 in any of the primary or secondary outcomes.

Exploratory analysis of efficacy of mirabegron in complete versus incomplete SCI
There were 30 patients with complete SCI, and 38 with incomplete SCI. The change in the PPBC met our exploratory p<0.25 threshold. Patients with complete SCI (ASIA A) improved by 0.4 points, however patients with incomplete (ASIA B, C, D E) improved by 1.2 points (p=0.09).

Exploratory analysis of efficacy of mirabegron in high versus low SCI
There were 38 patients with high SCI (≥T6), and 29 with low SCI (≤T7). The primary outcomes did not meet our exploratory p<0.25 threshold, however, patients with high SCI had a greater reduction in max detrusor pressure (-24.1mL) compared to those with low SCI (-3.1mL, p=0.21) and patient with low SCI had a greater improvement in their IQOL score (15.4) versus high SCI (7.8, p=0.24).

We have conducted an IPDMA on the two available placebo controlled randomised trials of mirabegron in people with MS or SCI. Using this dataset, we demonstrated a significant improvement in people treated with mirabegron compared to placebo in both urodynamic values (such as capacity, peak detrusor/NDO pressure), patient reported outcome measures (PPBC and I-QoL scores), and incontinence pad weight. A systematic review of OAB anticholinergic efficacy in NLUTD provides a reference against which to compare mirabegron; the improvement in bladder capacity (+49mL anticholinergics versus +49mL with mirabegron) was similar, however the decrease in maximum detrusor pressure was more divergent (-38cmH2O anticholinergics versus -18cmH2O with mirabegron). 

This analysis confirms that maximum detrusor pressure is significantly reduced with the use of mirabegron; this was not significantly different in the individual study analyses, likely as a result of the smaller sample size. This is obviously an important parameter that is linked to the risk of renal deterioration. The improvement in all urodynamic parameters that we evaluated (other than compliance), challenge the conclusion of prior systematic reviews which have stated that mirabegron does not improve urodynamic parameters in this patient population and metanalyses which have not shown significant results.  Our exploratory analysis has confirmed that mirabegron has good efficacy in both people with MS, and those with SCI. To our knowledge the impact of SCI lesion completeness and lesion level haven’t been explored in other analyses of oral medication for NLUTD. While there were some potential signals that our hypotheses that mirabegron has better efficacy in people with lower SCI, and incomplete lesions may be valid, prospective study of these groups is necessary before any conclusions can be drawn.

Compared to placebo, mirabegron significantly improves both urodynamic values (such as capacity, peak detrusor/NDO pressure), patient reported outcome measures (PPBC and I-QoL scores), and amount of incontinence. Further studies are needed to look at whether some groups of people with NLUTD achieve better outcomes with mirabegron or anticholinergic medications.

Krhut J, Borovicka V, Bílková K, Sýkora R, Míka D, Mokriš J, et al. Efficacy and safety of mirabegron for the treatment of neurogenic detrusor overactivity—Prospective, randomized, double-blind, placebo-controlled study. Neurourol Urodynam. 2018 Mar 31;37(7):2226–33.
Welk B, Hickling D, McKibbon M, Radomski S, Ethans K. A pilot randomized-controlled trial of the urodynamic efficacy of mirabegron for patients with neurogenic lower urinary tract dysfunction. Neurourol Urodynam. 2018 Aug 31;37(8):2810–7.