56
The mean percentage of paresthesia coverage for burst is significantly lower than that for tonic.
The following table summarizes the results of the test of the hypothesis of equality between burst
and tonic stimulation. On average, subjects reported only 4.5% of the mapped body segments
had paresthesia with burst stimulation compared to 22.7% with tonic stimulation. This difference
represents an 80.2% relative reduction in paresthesia coverage with burst stimulation over tonic
stimulation.
Table 32. Percentage of paresthesia coverage
Randomization
Arm 1:
tonic/burst
Arm 2:
burst/tonic
Pooled
Burst
Mean ± SD (N)
3.9±6.6 (37)
5.0±10.5 (36)
4.5±8.7 (73)
Minimum, Median, Maximum
0.0, 0.0, 23.4
0.0, 0.0, 51.1
—
Tonic
Mean ± SD (N)
20.1±14.3 (37) 25.3±18.1 (36) 22.7±16.3 (73)
Minimum, Median, Maximum
0.0, 17.0, 59.6
0.0, 20.2, 66.0
—
Burst-Tonic
Mean ± SD (N)
-16.2±14.5 (37) -20.3±19.5 (36)
—
Minimum, Median, Maximum
-57.4, -12.8, 8.5
-61.7, -18.1,
12.8
—
Burst-Tonic Average ± Pooled SD
Across Arms
-18.2±8.6
95% CI
22.2, -14.2
Additional Nonpowered Effectiveness
The following data consist of descriptive secondary and additional endpoints that were
predetermined within the clinical study protocol. See "Clinical Endpoints" (page 34). Additional
descriptive (nonpowered) data have also been provided.
Pain Measures
The FDA requested analyses to be performed on overall pain, trunk and limb VAS pain measures
from baseline to 12 weeks. Because the study had a crossover design in which subjects were
randomized to receive either tonic stimulation or burst stimulation from activation to 12 weeks,
these analyses were performed to evaluate study treatment to exclude the possibility of an
unequal carryover effect that would influence the results. Within the crossover design of this
study, analyses of the data from baseline to 24 weeks would not provide the same meaningful
evaluation of the study treatment.
As shown in the following table, the average changes in overall VAS score from baseline to week
12 were -32.1 and -26.4 for burst and tonic stimulation, respectively. The difference was -5.8 with
a 95% UCB for a mean difference of 2.9, which is less than the non-inferiority margin of 7.5.
Therefore, despite the smaller sample size, the primary endpoint of non-inferiority was met.