A Pooled Analysis of Three Clinical Studies: RESULTS

Posted by James

Demographic Data

Table 1 shows a comparison of the demographic data from the Rowbotham and Rice studies and for the pooled analysis.

Only patients who had at least 1,800 mg/day of gabapentin or placebo at baseline and at the end of study were included in the pooled analysis. The total number of patients from the three ITT populations was 334 + 229 + 40 = 603. A total of 603 ITT patients were analyzed. There were more women (54.5%, 329/603) than men (45.4%, 274/603). Most patients were 65 years of age or older (78.5%, 473/603), and almost 50% were older than 75 (46.8°%, 282/603). The median age was 74.

It is noteworthy that in both pivotal trials, the number of patients who discontinued gabapentin treatment (20%) did not differ significantly from the number who discontinued placebo treatment (15%).

Table 1 Baseline Characteristics of Patients with Postherpetic Neuropathic Pain

	Gabapentin (n = 358)	Placebo (n = 245)
Pivotal Studies
8-week trial
(Rowbotham et al., 1998)
Sex, No. (%)
Men	62 (56.9)	56 (48.3)
Women	47 (43.l)	60 (5l.7)
Age (years)
Mean
Median	73	74
Range	40-90	39-89
7-week trial
(Rice, 2001)
Sex, No. (%)
Men	92 (41.2)	46 (4l.l)
Women	l3l (58.7)	65 (58.6)
Age (years)
Mean	73.7	72.8
Median	75.5	74.9
Range	(22.5-90.8)	(28.9-94.8)
Pooled Analysis
Sex, No. (%)
Men	l62 (45.3)	ll2 (45.7)
Women	l96 (54.7)	l33 (54.3)
Age (years)
Mean	72.3	72.3
Median	74	74
Range	(22 – 90)	(28 – 94)
Age groups, n (%)
<65	77 (2l.5)	53 (2l.6)
>65	28l (78.5)	l93 (78.4)
>75	l70 (47.5)	ll2 (45.7)

Pooled Analysis

Gabapentin significantly improved overall mean pain scores in the pooled analysis of PHN patients, as measured by the Likert scale and the VAS, compared to those taking placebo (P = .0001). These findings are consistent with previous results, that gabapentin significantly decreased Likert scores and VAS scores in both PHN pivotal trials, and they reinforce the efficacy of gabapentin in the treatment of PHN.
canadian cialis online

In addition to its effect on pain, gabapentin also produced improvements in six of eight SF-36 subscale domains in PHN patients compared with patients receiving placebo (Figure 1). Consistent with results from both of the key trials, gabapentin significantly improved bodily pain (P = .0001), vitality (P = .0001), and mental health (P = .0001) domains.

Figure 1    Results of a pooled analysis, in which gabapentin produced significant improvements in six of eight Medical Outcomes Survey Short Form (SF)-36 subscale domains compared to placebo. The statistical analysis (analysis of covariance [ANCOVA]) was based on patient responses to items on the SF-36 questionnaire. Eight domains were assessed: bodily pain, vitality, mental health, role-physical, physical functioning, general health, social functioning, and role-emotional. ANCOVA models included baseline (the covariance), the study site, and treatment.

Physical functioning also improved measurably in the gabapentin patients (P = .03), role-physical (P = .02), and social functioning (P = .04) domains (Figure 1). The numbers of patients varied slightly between SF-36 domains because of missing measurements. Baseline values were equivalent between the two groups. Endpoint mean scores for gabapentin versus placebo differed by five or more points in four of the eight SF-36 subscale domains: bodily pain (9.212), vitality (8.036), mental health (5.392), and role-physical (7.312).
buy antibiotics online without prescription

Table 2 Reduction in Overall Mean Scores in Patients with Postherpetic Neuralgia Pain After Treatment with Gabapentin or Placebo

	Reduction in Likert Scale*	Visual Analogue Scalej
Placebo	-0.8	-12.38
Gabapentin	-2.21	-23.49
P value	P = .0001	P < .0001
* 0 = no pain, 10 = worst pain. j 0 mm = no pain, 100 mm = worst possible pain. Data from Pfizer, New York.

Table 2 includes the norms for the SF-36 domains for the general population. Baseline SF-36 scores for PHN patients in the pooled analysis were about one third lower than those of the general population, indicating a substantial diminution in health status among PHN patients. Their scores were also lower than patients who were 55 to 74 years of age, the group that encompasses the mean and median ages of the study population.

In the pooled analysis, PGI-C scores were assessed in 540 patients (gabapentin, n = 318; placebo, n = 222) and CGI-C scores were assessed in 548 patients (gabapentin, n = 324; placebo, n = 224). Patients receiving gabapentin reported overall relief of symptoms twice as frequently (65.4%, 208/318) than patients receiving placebo (34.7%, 77/222). The percentage of patients who reported that their pain was “very much improved” or “much improved” after taking gabapentin was 44.7% (142/318); in contrast, the percentage of patients reporting improvement with placebo was 18.5% (41/222).

Clinicians’ global impressions also showed reported “overall relief” of symptoms twice as frequently in the gabapentin patients (66.0°%, 214/324) than in the placebo patients (35.7°%, 80/224). The clinicians said that pain was “very much improved” or “much improved” in 45.1% of gabapentin patients (146/324) compared with 17% of patients taking placebo (38/224) (Table 3). These findings are consistent in both pivotal studies in which gabapentin medication provided notable pain relief, as assessed by both patient and CGI-C Questionnaires.