Wisconsin Ginseng to Improve Cancer-Related Fatigue
Wisconsin Ginseng to Improve Cancer-Related Fatigue
Between October 2008 and July 2011, 364 patients were enrolled from 40 different sites, mostly community cancer centers. All analyses were based on data frozen on November 29, 2011. A CONSORT diagram is shown as Figure 1. Seventy-eight percent of the participants completed all study interventions. Patient characteristics of those who started the study treatment were well balanced between arms (Table 1). There were no statistically significant differences in any of the fatigue measures between arms at baseline. All MFSI subscales are shown in Table 1. There were also not any statistically significant differences in demographic characteristics or fatigue scores between those who cancelled their participation after consent and randomization but before beginning the study treatment (cancellations) and those who began the study drug. Spearman correlation coefficients for the general fatigue subscale and pain were 0.08 (P = .13) and 0.01 (P = .80) between fatigue and sleep, respectively.
(Enlarge Image)
Figure 1.
CONSORT diagram. AE = adverse event.
The primary endpoint of change from baseline in the general subscale of the MFSI-SF at 4 weeks was 14.4 (standard deviation [SD] = 27.1) in the ginseng arm (n = 147) and 8.2 (SD = 24.8) in the placebo arm (n = 153) (P = .07). At 8 weeks, there was statistically significant improvement in fatigue for those on ginseng (n = 138) vs those on placebo (n = 133), with change scores of 20 (SD = 27) vs 10.3 (SD = 26.1), respectively (P = .003) (Figure 2). The magnitude in response for the group as a whole is shown in Figure 3. More participants had a positive response to the ginseng and more had a strong clinical benefit (≥30% improvement) from ginseng compared with placebo.
(Enlarge Image)
Figure 2.
Multidimensional Fatigue Symptom Inventory–Short Form general fatigue subscale change from baseline at 4 and 8 weeks. Differences between arms at 4 and 8 weeks were carried out with a two-sample, two-sided t test. All statistical tests were two-sided.
(Enlarge Image)
Figure 3.
Eight-week response to ginseng vs placebo (percentage of participants) per Multidimensional Fatigue Symptom Inventory–Short Form general subscale determined by χ test. All statistical tests were two-sided.
Secondary endpoints included the other subscales of the MFSI-SF, as well as the fatigue-inertia and vigor-activity subscales of the POMS. Change from baseline for these is shown in Table 2. Statistically significant improvements in fatigue were reported in the ginseng group over the placebo group for the physical subscale and total score of the MFSI-SF and the fatigue/inertia subscale of the POMS. The BFI total score and activity interference did not demonstrate statistically significant differences between the arms; however, individual items of worst fatigue and fatigue "now" were significantly different at 8 weeks, favoring the ginseng arm (data not shown).
When the population was divided into two groups, those receiving cancer treatment vs those who had completed treatment, and the primary analysis was repeated within each of those groups, participants undergoing cancer therapy assigned to the ginseng arm had statistically significant improvement in fatigue at 4 and 8 weeks compared with those in the placebo arm (Figure 4).
(Enlarge Image)
Figure 4.
Percentage change from baseline for general subscale of Multidimensional Fatigue Symptom Inventory–Short Form at 4 and 8 weeks by current vs postcancer treatment determined by χ test. All statistical tests were two-sided. The 4 week data were analyzed with a Wilcoxon signed rank test, and the 8 week data were analyzed with equal variance t-tests.
Only five toxicities greater than 1% incidence were attributed to study treatment, and these were not statistically significantly different between arms per CTCAE grading by study personnel. These toxicities were agitation, anxiety, insomnia, nausea, and vomiting (Table 3). Patient-reported toxicities, controlling for baseline, were also not statistically significantly different between the arms, over the 8 weeks of treatment. Scores changed little over the course of the study (no more than 5 points out of 100) for nausea, vomiting, nervousness, anxiety, trouble sleeping, and loose stools. Only loose stools at 4 weeks (–0.8) and pain at 8 weeks (–0.3) were worse than baseline, and these occurred only in the placebo group. All other symptoms improved over the course of the study.
Results
Between October 2008 and July 2011, 364 patients were enrolled from 40 different sites, mostly community cancer centers. All analyses were based on data frozen on November 29, 2011. A CONSORT diagram is shown as Figure 1. Seventy-eight percent of the participants completed all study interventions. Patient characteristics of those who started the study treatment were well balanced between arms (Table 1). There were no statistically significant differences in any of the fatigue measures between arms at baseline. All MFSI subscales are shown in Table 1. There were also not any statistically significant differences in demographic characteristics or fatigue scores between those who cancelled their participation after consent and randomization but before beginning the study treatment (cancellations) and those who began the study drug. Spearman correlation coefficients for the general fatigue subscale and pain were 0.08 (P = .13) and 0.01 (P = .80) between fatigue and sleep, respectively.
(Enlarge Image)
Figure 1.
CONSORT diagram. AE = adverse event.
The primary endpoint of change from baseline in the general subscale of the MFSI-SF at 4 weeks was 14.4 (standard deviation [SD] = 27.1) in the ginseng arm (n = 147) and 8.2 (SD = 24.8) in the placebo arm (n = 153) (P = .07). At 8 weeks, there was statistically significant improvement in fatigue for those on ginseng (n = 138) vs those on placebo (n = 133), with change scores of 20 (SD = 27) vs 10.3 (SD = 26.1), respectively (P = .003) (Figure 2). The magnitude in response for the group as a whole is shown in Figure 3. More participants had a positive response to the ginseng and more had a strong clinical benefit (≥30% improvement) from ginseng compared with placebo.
(Enlarge Image)
Figure 2.
Multidimensional Fatigue Symptom Inventory–Short Form general fatigue subscale change from baseline at 4 and 8 weeks. Differences between arms at 4 and 8 weeks were carried out with a two-sample, two-sided t test. All statistical tests were two-sided.
(Enlarge Image)
Figure 3.
Eight-week response to ginseng vs placebo (percentage of participants) per Multidimensional Fatigue Symptom Inventory–Short Form general subscale determined by χ test. All statistical tests were two-sided.
Secondary endpoints included the other subscales of the MFSI-SF, as well as the fatigue-inertia and vigor-activity subscales of the POMS. Change from baseline for these is shown in Table 2. Statistically significant improvements in fatigue were reported in the ginseng group over the placebo group for the physical subscale and total score of the MFSI-SF and the fatigue/inertia subscale of the POMS. The BFI total score and activity interference did not demonstrate statistically significant differences between the arms; however, individual items of worst fatigue and fatigue "now" were significantly different at 8 weeks, favoring the ginseng arm (data not shown).
When the population was divided into two groups, those receiving cancer treatment vs those who had completed treatment, and the primary analysis was repeated within each of those groups, participants undergoing cancer therapy assigned to the ginseng arm had statistically significant improvement in fatigue at 4 and 8 weeks compared with those in the placebo arm (Figure 4).
(Enlarge Image)
Figure 4.
Percentage change from baseline for general subscale of Multidimensional Fatigue Symptom Inventory–Short Form at 4 and 8 weeks by current vs postcancer treatment determined by χ test. All statistical tests were two-sided. The 4 week data were analyzed with a Wilcoxon signed rank test, and the 8 week data were analyzed with equal variance t-tests.
Toxicities/Side Effects
Only five toxicities greater than 1% incidence were attributed to study treatment, and these were not statistically significantly different between arms per CTCAE grading by study personnel. These toxicities were agitation, anxiety, insomnia, nausea, and vomiting (Table 3). Patient-reported toxicities, controlling for baseline, were also not statistically significantly different between the arms, over the 8 weeks of treatment. Scores changed little over the course of the study (no more than 5 points out of 100) for nausea, vomiting, nervousness, anxiety, trouble sleeping, and loose stools. Only loose stools at 4 weeks (–0.8) and pain at 8 weeks (–0.3) were worse than baseline, and these occurred only in the placebo group. All other symptoms improved over the course of the study.