Sodium oxybate relieves pain and improves function in fibromyalgia syndrome: A randomized, double-blind, placebo-controlled, multicenter clinical trial

Abstract

Objective

To evaluate the safety and efficacy of sodium oxybate for management of the symptoms of fibromyalgia syndrome (FMS).

Methods

Patients with FMS (according to the American College of Rheumatology 1990 criteria) were randomized, after discontinuing their prestudy medications for FMS, to receive 4.5 gm or 6 gm of sodium oxybate or matching placebo once per night for 8 weeks. The primary outcome variable (POV) was a composite score for changes from baseline in 3 coprimary self-report measures: patient's pain rating (in daily electronic diaries) on a visual analog scale (PVAS), the Fibromyalgia Impact Questionnaire (FIQ) score, and the Patient Global Impression of Change (PGI-C). A beneficial response rate for the POV composite score was defined as ≥20% improvement in the PVAS and FIQ scores plus a rating of “much better” or “very much better” on the PGI-C. Secondary measures included subjective sleep outcomes (on the Jenkins Scale for Sleep) and quality-of-life measures. The analyses were based on an intent-to-treat (ITT) population.

Results

The ITT population included 188 patients with FMS, 78% of whom completed the trial. Significant benefit was observed with both dosages of sodium oxybate, according to changes in the POV and subjective sleep quality. Improvements in the PVAS score were significantly correlated with sleep outcomes. Sodium oxybate was well tolerated overall; dose-related nausea (≤28% of patients) and dizziness (≤18% of patients) tended to resolve with continued therapy.

Conclusion

Sodium oxybate therapy was well tolerated and significantly improved the symptoms of FMS. Further study of sodium oxybate as a novel therapeutic option for FMS is warranted.

Fibromyalgia syndrome (FMS) is characterized by chronic widespread pain and allodynia in at least 11 of 18 anatomically defined tender points (1). Although it is most common in adult women, FMS has a prevalence of ∼2% in the general population, affecting an estimated 3–6 million people in the US (2). In 1997, estimated annual direct medical costs per patient totaled more than $2,000, with an annual cost to the US economy exceeding $12 billion (3).

Research classification criteria for FMS were endorsed by the American College of Rheumatology (ACR) in 1990 (1). These criteria have been used in the clinical arena for the diagnosis of FMS. A clinical case definition has been proposed as a consensus document (4). Although FMS is classified by the presence of pain and tenderness, many other clinical features have been identified in patient subgroups (4–6). Such features include insomnia, fatigue, stiffness, headaches, cognitive dysfunction, affective symptoms, and symptoms of irritable bowel syndrome.

The pathogenesis of FMS seems to involve a genetic predisposition, an initiating stimulus, dysregulation of neurotransmitter function, and central sensitization (7–10). The prevalence of both anxiety and depression in patients with FMS is higher than that in the normal population (11), but psychiatric comorbidity is not necessary for the pathogenesis of FMS (12).

The management of FMS is currently symptom-based, with the focus on alleviating pain, increasing restorative sleep, and improving physical function. Nonpharmacologic therapies include education, cognitive-behavioral therapy, sleep-hygiene modification, and exercise (13). Pharmacologic treatments are aimed at correcting biologically relevant abnormalities in FMS, using neuromodulatory, analgesic and/or sedative-hypnotic agents to address the pain and insomnia. No single drug has yet to demonstrate efficacy for all manifestations. To date, only one drug, pregabalin, has been approved with an indication specifically for the treatment of FMS. There is a need to expand the pharmacologic options.

Sodium oxybate is the sodium salt of γ-hydroxybutyrate (GHB), a metabolite of γ-aminobutyric acid. It is approved by the US Food and Drug Administration (FDA) for the treatment of cataplexy and excessive daytime sleepiness in patients with narcolepsy. In a clinical study involving patients with narcolepsy, sodium oxybate was shown to consolidate fragmented sleep, increase slow-wave sleep, and reduce nighttime awakenings (14).

In early open-label trials of sodium oxybate in patients with narcolepsy, some patients with coexisting FMS showed improvement in their FMS-related symptoms, in addition to improvement in their narcolepsy symptoms. Thus, Scharf and colleagues (15) predicted that sodium oxybate might benefit patients with FMS-related insomnia. In a small, randomized, placebo-controlled trial of sodium oxybate in patients with FMS (16), treatment with sodium oxybate significantly increased total sleep time, enhanced slow-wave sleep, reduced high-frequency alpha-wave intrusion into slow-wave sleep, reduced the frequency of nighttime awakenings, and reduced daytime fatigue. It also decreased the severity of pain and the tender point index (TPI).

Sodium oxybate for therapeutic use in narcolepsy is classified as a controlled substance under schedule III of the US Controlled Substances Act of 1970 and is dispersed by a single-distribution pharmacy. Nonmedical, illicitly used sodium oxybate and GHB are classified under schedule I, the most restrictive schedule of the Controlled Substances Act. In 2000, the US Drug Enforcement Agency added GHB and its analogs to the controlled substance schedules. The 2005 annual reports from the Drug Abuse Warning Network and the American Association of Poison Control Centers indicated that there have been significant decreases in exposures to GHB or GHB-related analogs (17, 18). Although there have been no reports of dependence or withdrawal associated with the use of sodium oxybate in clinical trials in patients with narcolepsy, there have been case reports of physical dependence after illicit use of GHB, when taken in frequent, repeated doses in excess of the therapeutic dosage range (18–250 gm/day).

The objectives of this proof-of-concept, 8-week, randomized, double-blind, placebo-controlled, parallel-group, dose-ranging study were to determine whether sodium oxybate would be more effective than placebo in reducing the severity of pain, dysfunction, and insomnia in patients with FMS, and to assess the safety and side-effect profile of sodium oxybate in patients with FMS.

PATIENTS AND METHODS

Oversight.

The study was conducted in accordance with the Declaration of Helsinki (as described on the US FDA Web site at http://www.fda.gov/oc/health/helsinki89.html) and received approval from the institutional review board at each site. The study sponsor monitored compliance according to good clinical practice guidelines (E6), based on the International Conference on Harmonisation Guidelines for Good Clinical Practice in the European Community.

Patients.

Men or women ages ≥18 years who met the ACR 1990 classification criteria for FMS (1) were recruited and enrolled in the study. All patients gave their written informed consent.

Inclusion criteria required that the patient's mean score on the visual analog scale (VAS) for pain (PVAS) be >40 on a 0–100-point VAS, as recorded in the patient's diary for up to 2 weeks before randomization at visit 4 (representing the baseline PVAS score). In addition, the patients agreed to adhere to the following rules during the study: required discontinuation of opiates, antidepressants, cyclobenzaprine, and tramadol, allowed continuation of a preexisting regimen of acupuncture, physical therapy, and/or massage therapy, allowed continuation of all behavioral and cognitive therapies, use of rescue analgesic therapies restricted to only 1 over-the-counter medication (either acetaminophen ≤4,000 mg/day, ibuprofen ≤1,200 mg/day, naproxen ≤660 mg/day, or ketoprofen ≤75 mg/day), prohibited use of alcohol, and for those women who were not surgically sterile or not postmenopausal for at least 2 years prior to the study, required use of a medically accepted method of birth control.

Recruited study patients were excluded if any of the following were found to apply: presence of an inflammatory rheumatic disease, presence of a painful disorder other than FMS, presence of a medical or psychological condition that might compromise participation in the study, an Apnea Hypopnea Index of >15 per hour on a screening polysomnogram (exempted if receiving continuous positive airway pressure therapy), presence of a seizure disorder, a history of head trauma resulting in loss of consciousness, a history of migraine headaches, previous intracranial surgery, a current or recent history (within 1 year) of any substance abuse disorder (including abuse of alcohol), deficiency in succinic semialdehyde dehydrogenase, the use, in the 30 days prior to the screening visit, of certain medications (sodium oxybate, any investigational therapy, any anticonvulsant agent [even if willing to discontinue its use], any antidepressant [exempted if discontinued for at least 5 half-lives], sleep aids such as hypnotics, tranquilizers, and antihistamines [exempted if nonsedating antihistamines were taken], and benzodiazepines or clonidine), a serum creatinine level >2.0 mg/dl, abnormal results on liver function tests (transaminase levels at least twice the upper limit of normal or serum bilirubin levels at least 1.5 times the upper limit of normal), a positive pregnancy test result, an electrocardiogram that disclosed an arrhythmia or an atrioventricular conduction delay of more than first-degree block, pending litigation for worker's compensation or other monetary settlements, or an occupation that required night-shift work.

Clinical sites.

The study was conducted at 21 clinical sites in the continental US, by personnel experienced in diagnosing FMS and caring for patients with FMS. Collectively, the health-care providers at these sites were identified as the Oxybate SXB-26 FMS Study Group (see Acknowledgments for participating sites and investigators).

Study design and treatment.

Patients were screened for eligibility and were evaluated for sleep apnea. After undergoing a treatment withdrawal/washout period of up to 39 days if needed, patients recorded PVAS scores for up to 2 weeks as a baseline measure. Investigators completed the Clinical Global Impression of Severity (CGI-S) assessment at the end of the baseline period (visit 4). Patients who met the entry criteria were randomized to receive an oral solution of sodium oxybate (4.5 gm/night or 6 gm/night) or 1 of 2 placebo doses (matched to active treatment by volume) without dosage titration. The study medication was administered in a blinded manner for 8 weeks.

Therapeutic protocol.

Sodium oxybate has a half-life of 30–60 minutes. To maintain efficacy in controlling sleep disturbances throughout the night, the total daily dose of the blinded medication was equally divided between a bedtime dose and a second dose 2.5–4 hours later. To help the study patients with compliance, they were advised to set an alarm as a reminder for the timing of the second dose. The mean number of tablets of rescue analgesic medication was calculated per patient over a 14-day period.

Efficacy assessments.

Primary and secondary efficacy parameters were recorded in the daily electronic diaries and were assessed at clinic visits after 2, 4, and 8 weeks of treatment (visits 5, 6, and 7, respectively). Each morning, afternoon, and evening, patients used an electronic diary (PHT Corporation, Charlestown, MA) to rate 2 self-report outcome measures: the PVAS score (0 = no pain, 100 = worst possible pain), and the fatigue VAS (FVAS; 0 = no fatigue, 100 = worst possible fatigue) (19, 20).

During clinical assessment visits, patients completed the following questionnaires: the Fibromyalgia Impact Questionnaire (FIQ) (19), the Jenkins Scale for Sleep (JSS) (21), and a quality-of-life questionnaire (the Short Form 36 [SF-36] Health Survey) (22); the ranges of the FIQ and SF-36 are 0–100 and that of the JSS is 0–20, with higher scores indicating greater severity. At the final visit, patients completed the Patient Global Impression of Change (PGI-C) assessment (23). Depression was a criterion for exclusion from randomization, but self-reported affective symptoms were assessed in both the FIQ and the SF-36.

Assessments completed by the clinician at each of the clinic visits included measuring the tender point count (TPC; range 0–18) and TPI (based on calibrated digital deep pressure; range for each tender point 0–4) (24), the CGI-S at baseline, and the Clinical Global Impression of Change (CGI-C) at the end of the study. Polysomnography was performed at visits 2, 4, 6, and 7, but the outcomes from those assessments are being reported separately (25).

Tolerability and safety assessments.

All spontaneously reported or observed treatment-emergent adverse events (AEs) were recorded at each clinic visit, along with the dates of onset and resolution of the AEs. A 12-lead electrocardiogram was recorded at the time of screening and at the study end point. Clinical laboratory tests (hematology, chemistry, and urinalysis) were performed at visits 1, 4 and 7. Any serious or unexpected AEs that occurred during the course of the study were reported within 24 hours of occurrence to the sponsor and to each institutional review board.

Statistical analysis.

Randomization was achieved by computer-generated coding, using a block size of 6. To be counted in the intent-to-treat (ITT) analysis, patients who were randomized to a treatment group were required to have valid baseline data for the PVAS and FIQ components of the primary outcome variable (POV) and to have received at least 1 dose of blinded study medication (sodium oxybate at a dosage of 4.5 gm/night or 6 gm/night, or placebo).

Consistent with the recommendations of the US FDA Arthritis Advisory Committee (from June 23, 2003), the POV used in the study was a composite score that required specified levels of improvement in 3 measures for a patient to be considered a responder. The 3 components of the POV were 1) the PVAS score from the electronic diaries (responders were required to achieve ≥20% reduction in the mean PVAS score when comparing the week before visit 7 with the week before baseline), 2) the FIQ score (responders were required to achieve ≥20% reduction in the FIQ score from baseline [visit 4] to visit 7), and 3) the PGI-C (responders were those who had a PGI-C rating of “much better” or “very much better” at visit 7). In addition, ≥30% and ≥50% POV responder analyses were performed, in which responders were required to show a reduction of at least 30% or at least 50% in the PVAS score and FIQ score and to provide a PGI-C rating of “much better” or “very much better” at visit 7. Secondary outcome measures included each of the POV components examined separately, as well as changes in the FVAS score, JSS score, SF-36 score, TPC/TPI, and CGI-C rating, as well as use of rescue medications.

Safety measures included the analysis of AEs, clinical laboratory values, electrocardiograms, vital statistics, and physical examination findings. Efficacy was assessed in the ITT population. For continuously distributed outcomes, the active treatment groups (sodium oxybate 4.5 gm/night or 6 gm/night) were compared with the placebo group, in terms of comparing the mean change from baseline to study end point. For binary analyses of POV responders, patients with incomplete followup data were defined as nonresponders. For other measures, the last observation carried forward (from baseline) was used to impute missing data from the baseline value. Comparisons of treatment groups were carried out with analysis of covariance, adjusted for site and treatment group–by-site interactions; if the P value for the interaction was >0.10, then the term was dropped from the model.

If the model assumptions were violated, Wilcoxon's test was used. For binary end points, Fisher's exact test was used to compare the proportion of responders between the active treatment groups and the placebo group. Safety was assessed in all patients who received at least 1 dose of the test article (sodium oxybate or placebo), and missing values were not imputed. For other analyses, least squares mean values and their standard errors were determined from a repeated-measures linear model with an autocorrelation matrix in autoregressive order one. The mean number of doses of rescue pain medication taken in the 2-week period before the study end point was compared with the mean number of doses taken at baseline (the 2 weeks prior to randomization).

All statistical testing was 2-sided with a significance level of 5%. SAS for Windows version 9.1 (SAS Institute, Cary, NC) was used for all analyses.

RESULTS

General characteristics of the patients in the randomized and withdrawal groups.

Of the 320 patients screened, 195 were randomized to receive a treatment (Figure 1). Of those, 188 were included in the ITT analysis because they took at least 1 dose of blinded study medication and had complete baseline data at visit 4. The majority of recruited patients who were excluded by the screening procedures had failed to discontinue prestudy medications.

Figure 1. Flow diagram of enrollment, randomization, withdrawal, and study completion per protocol (PP) among patients with primary fibromyalgia syndrome receiving treatment with either placebo or sodium oxybate at a dosage of 4.5 gm/night or 6 gm/night. AE = adverse event.

The baseline characteristics of the ITT population were similar across treatment groups (Table 1). The majority of patients were female and white. The mean duration of FMS was ∼9 years, and in 61% of patients FMS was diagnosed more than 5 years prior to the study start. The prevalence of irritable bowel syndrome was 30%.

Table 1. Baseline demographic characteristics of the intent-to-treat population of patients with fibromyalgia syndrome (FMS) (n = 188)

	Placebo	Sodium oxybate
		4.5 gm	6 gm
* Except where indicated otherwise, values are the mean ± SD. BMI = body mass index; PVAS = visual analog scale for pain.
No. of patients	64	58	66
Age, years	47.3 ± 10.6	47.4 ± 12.1	45.5 ± 11.6
White, %	89.1	89.7	97.0
No. female/no. male	61/3	54/4	63/3
Height, cm	162.7 ± 7.0	162.2 ± 7.8	164 ± 9.7
Weight, kg	77.7 ± 17.1	81.3 ± 21.5	82.9 ± 19.8
BMI, kg/m^2	29.4 ± 6.9	31.1 ± 9.1	30.8 ± 7.4
Duration of FMS >5 years, %	81.3	58.6	63.6
Insomnia, %	92.2	94.5	87.7
Initial insomnia >20 minutes, %	68.8	77.2	74.2
Slept <6 hours/night, %	81.3	75.9	62.1
Awoke >3 times/night, %	66.1	70.9	62.5
Awake >30 minutes per episode, %	79	73.2	64.1
PVAS score, mm	65.9 ± 15.6	65.8 ± 17.2	65.7 ± 17.3
Tender point count, median (range)	18 (12–18)	18 (13–18)	18 (12–18)
Tender point index, median (range)	40 (14–72)	35 (15–69)	39 (16–72)

In total, 147 patients (78.2%) completed the study per protocol (Figure 1). AEs led to early termination in 3 of 64 patients (4.7%) from the placebo group, 5 of 58 patients (8.6%) from the sodium oxybate 4.5 gm/night group, and 11 of 66 patients (16.7%) from the sodium oxybate 6 gm/night group.

Compliance with dosing schedule.

Typically, patients did not object to the timing of the second dose, because they had experienced late-night wakefulness for many months (up to years). Compliance with the use of the blinded medication was documented by comparing the prescribed volume with the difference between the dispensed and returned volumes. When the level of compliance was defined as having completed use of 70% of the prescribed medication, the mean proportion of compliant patients was 81% (Table 2).

Table 2. Changes in outcome variables from baseline to end point in patients with fibromyalgia syndrome after blinded treatment with 4.5 gm or 6 gm of sodium oxybate compared with placebo*

Outcome	Placebo	Sodium oxybate
		4.5 gm	P vs. placebo	6 gm	P vs. placebo
* Except where indicated otherwise, values are the mean ± SD. Patients were assessed on components of the primary outcome variable (POV) composite score, i.e., the pain rating on a visual analog scale (PVAS; scale 0–100), the Fibromyalgia Impact Questionnaire (FIQ) score (scale 0–100; subscale 0–10), the Patient Global Impression of Change (PGI-C), and the Clinical Global Impression of Change (CGI-C), and responders were defined as those achieving a ≥20% (or ≥30% or ≥50%) reduction in the mean PVAS score from baseline (visit 4) to visit 7, a ≥20% (or ≥30% or ≥50%) reduction in the mean FIQ score from baseline (visit 4) to visit 7, and a rating of “much better” or “very much better” on the PGI-C and “much improved” or “very much improved” on the CGI-C. Secondary outcome measures included a fatigue VAS (FVAS; scale 0–100) score, a Jenkins Scale for Sleep (JSS) score (scale 0–20), a Short Form 36 (SF-36) Health Survey score (scale 0–100), and the tender point count (TPC; scale 0–18) and tender point index (TPI; scale 0–4 per tender point). Except where indicated otherwise, all P values were determined by Fisher's exact test. The frequency of POV responders at the ≥20% level in the sodium oxybate groups did not differ significantly from that of POV responders at the ≥30% level in the sodium oxybate groups (P = 0.44 between 20% responders; P = 0.55 between 30% responders). NA = not applicable. † By analysis of variance, adjusted for site or on ranks adjusted for site.
Sample size	64	58	–	66	–
No. (%) with 70% compliance with prescribed dosage	54 (84)	51 (88)	NA	48 (73)	NA
No. (%) of POV responders
20% criteria	8 (12.5)	20 (34.5)	0.005	18 (27.3)	0.048
30% criteria	7 (10.9)	18 (31.0)	0.007	17 (25.8)	0.04
PVAS
Change in score from baseline	−8.6 ± 20.4	−16.2 ± 21.5	0.04†	−15.9 ± 20.2	0.03†
No. (%) of responders
20% criteria	15 (23.4)	28 (48.3)	0.005	30 (45.5)	0.01
30% criteria	13 (20.3)	24 (41.4)	0.02	26 (39.4)	0.02
50% criteria	9 (14.1)	16 (27.6)	0.08	19 (28.8)	0.05
FIQ
Change in score from baseline	−10.4 ± 17.8	−20.4 ± 21.4	0.007†	−18.4 ± 19.3	0.02†
No. (%) of responders
20% criteria	17 (26.6)	31 (53.4)	0.003	35 (53.0)	0.002
30% criteria	14 (21.9)	27 (46.6)	0.007	31 (47.0)	0.003
Change in subscale score from baseline
Physical impairment	−0.23 ± 0.55	−0.52 ± 0.65	0.002†	−0.37 ± 0.55	0.09†
Days felt good	1 ± 2.1	2 ± 2.4	0.01†	1.8 ± 2.1	0.01†
Work missed	−0.6 ± 1.9	−0.8 ± 2.0	0.37†	−0.7 ± 1.4	0.54†
Job ability	−1.0 ± 2.2	−2.3 ± 2.8	0.008†	−2.0 ± 2.6	0.02†
Pain	−1.0 ± 2.2	−2.2 ± 2.8	0.02†	−2.0 ± 2.4	0.02†
Fatigue	−1.1 ± 2.3	−2.6 ± 2.9	0.002†	−2.4 ± 2.6	0.004†
Tired upon awakening	−1.3 ± 2.1	−2.9 ± 3.1	0.003†	−2.4 ± 2.6	0.02†
Stiffness	−0.9 ± 2.4	−2.3 ± 2.8	0.004†	−2.3 ± 2.7	0.002†
Anxiety	−1.0 ± 2.1	−1.2 ± 2.6	0.88†	−1.0 ± 2.8	0.82†
Depression	−1.0 ± 2.2	−1.3 ± 2.6	0.64†	−1.4 ± 2.1	0.21†
No. (%) of PGI-C responders	12 (18.8)	22 (37.9)	0.03	21 (31.8)	0.11
No. (%) of CGI-C responders	16 (25)	24 (42.1)	0.05	31 (47)	0.01
Change in TPC from baseline, median (range)	0 (−16 to 2)	−1 (−16 to 2)	0.06†	−1 (−18 to 4)	0.05†
Change in TPI from baseline	−6.9 ± 14.2	−10.9 ± 13.1	0.06†	−10.5 ± 14.6	0.09†
Change in JSS from baseline, median (range)	−1 (−16 to 5)	−5.5 (−20 to 3)	0.004†	−6 (−18 to 1)	0.003†
Change in FVAS score from baseline	−10 ± 20	−17.4 ± 22.4	0.05†	−16.8 ± 19.7	0.05†
Change in SF-36 score from baseline
Physical component	2.8 ± 7.2	5.6 ± 8.9	0.08†	5.4 ± 7.2	0.09†
Mental component	3.0 ± 10.8	5.9 ± 10.8	0.21†	5.0 ± 11.5	0.27†
Physical role domain	2.6 ± 7.8	5.4 ± 9.7	0.17†	5.8 ± 9.8	0.05†
Pain index domain	3.5 ± 7.8	7.4 ± 9.7	0.02†	8.0 ± 9.0	0.008†
Vitality index domain	5.0 ± 9.8	10.9 ± 12.5	0.01†	10.0 ± 12.0	0.02†

Efficacy.

As shown in Table 2, 20% responder rates for the POV composite score in both the sodium oxybate 4.5 gm/night group (20 of 58 patients [34.5%]; P = 0.005 versus placebo) and 6 gm/night group (18 of 66 patients [27.3%]; P < 0.05 versus placebo) were significantly higher than in the placebo group (8 of 64 patients [12.5%]) but not significantly different from each other (P = 0.44). The 30% responder rates for the POV composite score in both the sodium oxybate 4.5 gm/night group (18 of 58 patients [31.0%]; P = 0.007 versus placebo) and 6 gm/night group (17 of 66 patients [25.8%]; P = 0.04 versus placebo) were significantly higher than in the placebo group (7 of 64 patients [10.9%]) (Figure 2) but not significantly different from each other (P = 0.55).

Figure 2. Primary outcome variable (POV) 30% responder analysis. The POV is a composite score in which the 30% POV response was defined as ≥30% improvement in the mean self-reported pain score on a visual analog scale, ≥30% improvement in the Fibromyalgia Impact Questionnaire score, and change in patient's global health assessment rating to “much improved” or “very much improved.” All 3 variables had to meet these improvement criteria for the patients with fibromyalgia syndrome to be counted as a responder at this level on the POV composite measure. Based on these criteria, the response of the patients to either dose of sodium oxybate (Oxy) was statistically significantly different from that in the placebo group. Bars show the mean and SD.

Changes in each of the individual components of the POV separately met the criteria for a significant difference between treatment groups (Table 2). At visit 7, the mean decrease in the PVAS score was significantly greater in both the sodium oxybate 4.5 gm/night group (mean ± SD −16.2 ± 21.5; P = 0.04 versus placebo) and 6 gm/night group (−15.9 ± 20.2; P = 0.03 versus placebo) compared with the placebo group (−8.6 ± 20.4). Figure 3A shows that the pattern of greater reduction in the PVAS score in both sodium oxybate groups was consistent throughout the study. Responder analyses for the PVAS variable at 20%, 30%, and 50% improvement levels showed that the active treatment was consistently more effective than the placebo treatment (Table 2).

Figure 3. Change from baseline in the mean morning pain visual analog scale (VAS) score (A) and mean morning fatigue VAS score (B) during the weeks before visits 5, 6, and 7 for all patients with fibromyalgia syndrome meeting the intent-to-treat criteria (n = 188). Patients were segregated by treatment group (placebo, sodium oxybate 4.5 gm/night, or sodium oxybate 6 gm/night). Results are the least squares mean change (±1 SEM). The baseline observation carried forward method was used to impute missing data for those patients who did not complete the study. ∗ = P < 0.05; ∗∗∗ = P < 0.001 versus placebo.

Similarly, there were significant reductions in total scores on the FIQ component of the POV (−10.4 ± 17.8 in the placebo group versus −20.4 ± 21.4 in the sodium oxybate 4.5 gm/night group [P = 0.007 versus placebo] and −18.4 ± 19.3 in the 6 gm/night group [P = 0.02 versus placebo]). The numbers of patients whose FIQ scores improved by at least 20% or at least 30% were significantly higher for both doses of sodium oxybate compared with placebo (Table 2). The same differences were also observed for most of the FIQ subscale scores (Table 2), with the exception of scores for work missed, anxiety, and depression.

A clinically relevant, subjectively rated change in pain, the third component of the POV, was indicated if a patient gave a rating of “much better” or “very much better” on the PGI-C. Table 2 shows that significantly more patients receiving sodium oxybate 4.5 gm/night (22 of 58 patients [37.9%]) reached these levels of improvement on the PGI-C compared with placebo-treated patients (12 of 64 [18.8%]; P = 0.03). However, these levels of improvement on the PGI-C did not differ significantly between the patients receiving 6 gm/night sodium oxybate (21 of 66 patients [31.8%]) and the placebo group (P = 0.11). Among the PGI-C responders, the mean ± SD changes in the PVAS score were −33.2 ± 23.4 in the placebo group (n = 12), −35.6 ± 19.1 in the sodium oxybate 4.5 gm/night group (n = 22), and −34.6 ± 20.8 in the sodium oxybate 6 gm/night group (n = 21).

Tender point survey.

The median reduction in the TPC was not significantly different in the 4.5 gm/night sodium oxybate group compared with the placebo group (P = 0.06) but was significantly different in the 6 gm/night sodium oxybate group compared with the placebo group (P = 0.05) (Table 2). The changes in the TPI were not significantly different for either active treatment group compared with the placebo group.

Insomnia.

As shown in Table 1, a majority of the 188 patients in the ITT population at baseline reported sleeping poorly (92.2% in the placebo group, 94.5% in the sodium oxybate 4.5 gm/night group, and 87.7% in the sodium oxybate 6.0 gm/night group). A high percentage of patients had initial insomnia (taking at least 20 minutes to fall asleep), awakened at least 3 times during the night, could not readily return to sleep, and achieved fewer than 6 hours of sleep per night.

Table 2 shows that both of the active treatment groups experienced significant improvement in sleep, based on the JSS (P = 0.004 for sodium oxybate 4.5 gm/night versus placebo; P = 0.003 for sodium oxybate 6 gm/night versus placebo). The mean fatigue score decreased in both of the sodium oxybate–treated groups compared with the placebo group (P = 0.05). At visit 7, there was more improvement in the FVAS score from baseline in both sodium oxybate groups (−17.4 ± 22.4 in the sodium oxybate 4.5 gm/night group [P = 0.05 versus placebo]; −16.8 ± 19.7 in the sodium oxybate 6 gm/night group [P = 0.05 versus placebo]) than in the placebo group (−10 ± 20). These significant differences in improvement of the FVAS score were observed early in the study and continued throughout (Figure 3B).

Quality of life.

Scores on the SF-36 quality-of-life assessment showed significant improvement in the pain index domain and vitality index domain for both of the sodium oxybate treatment groups compared with the placebo group (in the sodium oxybate 4.5 gm/day group, improvement in pain index P = 0.02 versus placebo, improvement in vitality index P = 0.01 versus placebo; in the sodium oxybate 6.0 gm/day group, improvement in pain index P = 0.008 versus placebo, improvement in vitality index P = 0.02 versus placebo) (Table 2).

Affective symptoms.

The study protocol did not permit the inclusion of patients with FMS who had psychiatric disorders. However, among participating patients, the affective symptoms measured on the FIQ did not improve nor did they worsen with sodium oxybate therapy (anxiety FIQ scores, P = 0.93 in the sodium oxybate groups versus placebo [2 degrees of freedom]; depression FIQ scores, P = 0.45 in the sodium oxybate groups versus placebo [2 df]).

Clinician global impression of change.

A significantly greater proportion of patients in both sodium oxybate treatment groups compared with the placebo group had a CGI-C rating of “much improved” or “very much improved” from baseline to visit 7 (Table 2).

Use of rescue medication.

There was no significant difference between the placebo group and either sodium oxybate treatment group in the percentage of patients who discontinued their use of rescue medication prior to visit 7 (20.3% in the placebo group, 27.6% in the sodium oxybate 4.5 gm/night group [P = 0.68 versus placebo], and 24.2% in the sodium oxybate 6 gm/night group [P = 1.00 versus placebo]). Within the 14 days prior to visit 7, the median rescue medication dosage, expressed as a percentage of the allowed dosage, was numerically, but not statistically significantly, lower in both of the sodium oxybate–treated groups compared with the placebo group (8.0% of the allowed dosage in the placebo group, 2.1% in the sodium oxybate 4.5 gm/night group [P = 0.08 versus placebo], and 3.6% in the sodium oxybate 6 gm/night group [P = 0.23 versus placebo]).

Correlational analysis.

Correlations were investigated between hypothesis-driven pairs of variables (data not shown). There were significant correlations between improvements in the JSS sleep measure and improvements in several key quality-of-life measures, the PVAS score (r = 0.55), the FVAS score (r = 0.55), the morning stiffness score (r = 0.59), and the TPI (r = 0.52).

Tolerability and safety.

Treatment-emergent AEs were reported in the majority of patients in all 3 treatment groups. These AEs occurred in 60.0%, 68.3%, and 77.6% of patients in the placebo group, sodium oxybate 4.5 gm/night group, and sodium oxybate 6 gm/night group, respectively. The differences in these proportions were not statistically significant (P = 0.09). There were no clinically important changes in vital signs, laboratory measures, general examination findings, neurologic examination findings, or electrocardiograms.

Two patients experienced serious AEs (1 experienced a respiratory infection with an exacerbation of asthma, while the other patient, who had a previously undisclosed bipolar affective disorder, experienced a manic episode, an episode of hypertension, and an episode of tachycardia). None of these AEs was judged to be related to the study drug or to study procedures. No deaths were reported in this study.

Twenty-two patients (11.5%) discontinued the study because of nonserious AEs. These AEs occurred in 3 patients receiving placebo (chest pain and tachycardia in 1, anxiety and depression in 1, and agitation in 1), 6 patients receiving sodium oxybate 4.5 gm/night (1 each with pruritus/rash, palpitations/insomnia, anxiety/insomnia, anxiety/paranoia, insomnia/muscle spasm/nausea/edema, and depression), and 13 patients receiving sodium oxybate 6 gm/night (5 with nausea with or without vomiting, and 1 each with syncope, chest pain/palpitations, dizziness, depression/panic, depression/suicidal concerns, insomnia, nausea/vomiting/asthenia, and constipation/urinary incontinence).

Most of the observed AEs were mild or moderate in severity. Table 3 shows AEs that occurred at a frequency of higher than 2% and that differed significantly in frequency between the active and placebo treatment groups. Of these, the dose-dependent AEs were nausea, nervous system manifestations, dizziness, renal/urinary disorders, and urinary incontinence.

Table 3. Adverse events affecting ≥2% of the patients with fibromyalgia syndrome in the all-treated population (n = 192)*

Adverse event	Placebo (n = 65)	Sodium oxybate		P versus placebo
		4.5 gm (n = 60)	6 gm (n = 67)
* Values are the number (%) of patients. P values were determined by Fisher's exact test (2 degrees of freedom). † Includes dizziness, headache, paresthesia, and somnolence.
Nausea	6 (9.23)	10 (16.67)	19 (28.36)	0.02
Pain in extremity	0 (0)	4 (6.67)	0 (0)	0.009
Nervous system disorders†	9 (13.85)	16 (26.67)	27 (40.3)	0.003
Dizziness	2 (3.08)	5 (8.33)	12 (17.91)	0.01
Restlessness	0 (0)	3 (5)	0 (0)	0.03
Renal and urinary disorders	0 (0)	1 (1.67)	5 (7.46)	0.04
Urinary incontinence	0 (0)	0 (0)	4 (6.0)	0.004

Ten patients (5.2%) reported experiencing either anxiety or depression as AEs (4 [2.1%] reporting anxiety and 6 [3.1%] reporting depression). One of these 10 patients reported experiencing both anxiety and depression, and 1 of the 6 patients with depression (0.5%) expressed suicidal ideation. Nine of the 10 patients were taking 1 of the 2 dosages of sodium oxybate, and all of the patients reporting depression were removed from the study because of these AEs. No signs of dependence or withdrawal symptoms were seen during the 1-week postdose washout period.

DISCUSSION

In this study, the therapeutic efficacy and safety of 2 different dosages of sodium oxybate were evaluated over 8 weeks in patients with FMS. The patients recruited for this study met the ACR research criteria for the classification of FMS (1). Their baseline assessments were comparable with those reported in several published trials involving patients with FMS (26–29).

The composition of the primary efficacy measure (the POV, assessing pain, physical dysfunction, and overall global impression of change) represented one approach to meeting US FDA guidelines for assessing the therapeutic effect of a pharmaceutical agent on the “syndrome” of FMS. The study was based on a responder analysis model in which a specific level of improvement in each of 3 important clinical variables was required, rather than simply comparing the mean variable values between treatment groups. A previous controlled trial in FMS had shown that these levels of change could be achieved with sodium oxybate therapy (16). In this proof-of-concept study, not only did the POV composite meet significance criteria after treatment with either dosage of sodium oxybate, but each of the 3 variables improved significantly when analyzed separately. Significant benefit with sodium oxybate therapy was apparent as early as visit 5 (after 2 weeks of therapy). The 2 dosages of sodium oxybate were not associated with a dose-dependent effect, but future studies could explore lower and higher dosages.

A high correlation was found between the PGI-C rating and the change in pain rating on the PVAS. The subset of patients with the best PGI-C outcomes consistently exhibited changes of >30% in PVAS scores, which supports similar observations by Farrar and colleagues (23) with a different agent.

Although pain has been the primary domain defining FMS, the majority of patients (75–80%) also experience sleep disturbance, fatigue, and stiffness (4–6). Patients in each of the sodium oxybate–treated groups exhibited significant improvements in the JSS sleep-outcome measure. Other evidence of improved sleep with sodium oxybate therapy was reported in a parallel study (25). Sodium oxybate is known for its sedative effects and for its ability to induce deep, restorative sleep (16), so it was predicted that its sedative properties might lessen the symptoms of insomnia associated with FMS. In animal models, prolonged sleep deprivation can lead to a variety of morbid consequences, such as immune insufficiency, thyroid dysfunction, and alteration in regional cerebral blood flow (30). In healthy humans, short-term, selective deprivation of slow-wave sleep lowers the pain threshold (31–34). It was reasonable to hypothesize that chronic sleep deprivation in FMS may contribute etiologically to widespread body pain and other symptoms. The high correlation (r = 0.55) between improvements in pain (on the PVAS) and sleep (on the JSS) with sodium oxybate treatment supports this hypothesis but does not prove a cause-and-effect relationship. Each of the sodium oxybate treatment groups in this study demonstrated improvements in both the physical and the fatigue-related components of the FIQ, as well as improvements in the pain and vitality domains of the SF-36, suggesting that sodium oxybate therapy–induced improvements in pain and fatigue translate into better quality of life (22).

Despite the fact that this study excluded patients with major depression, symptoms of depression and anxiety were both reported as AEs in a small proportion of sodium oxybate–treated patients, and this finding resulted in some patients being removed from the study. In contrast, there was no significant mean change in the self-reported severity of depression on a VAS or anxiety on a VAS as subscale components of the FIQ (see Table 2).

Sodium oxybate was generally well tolerated by patients with FMS. The side-effect profile of sodium oxybate was similar to that previously reported in the treatment of narcolepsy (35). As expected, the most frequently reported AEs were dose-related central nervous system symptoms, such as nausea, vomiting, and dizziness. Although the majority of sodium oxybate–treated patients experienced at least 1 AE, most of these events were mild to moderate in severity and seldom resulted in withdrawal from therapy. Even dose-dependent AEs tended to resolve with continued therapy. The overall discontinuation rate due to AEs was low in this study. Somnolence, which occurred quite rapidly after dosing, was considered a therapeutic benefit rather than an AE. Daytime sleepiness was not expected, because dosing occurred at night, and the half-life of the drug is very short.

In this trial, a subgroup of patients did not respond to sodium oxybate treatment, which likely reflects the heterogeneity of the patient population (36). Many agents with pharmacologically different mechanisms have demonstrated effectiveness in subgroups of FMS patients, but no single agent has met the needs of all patients nor addressed the complete spectrum of FMS symptoms.

In summary, this study is the second prospective, randomized clinical trial to evaluate the potential role of sodium oxybate in the management of FMS. Treatment with sodium oxybate led to significant changes in the POV composite measure. It reduced the severity of pain, insomnia, and fatigue, while improving the quality of life. Therefore, this proof-of-concept study suggests that sodium oxybate is effective and safe for the management of FMS symptoms. Further investigation of this agent for the treatment of FMS is warranted.

AUTHOR CONTRIBUTIONS

Dr. Russell had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Russell.

Acquisition of data. Russell, Perkins.

Analysis and interpretation of data. Russell, Perkins, Michalek.

Manuscript preparation. Russell, Perkins, Michalek.

Statistical analysis. Michalek.

ROLE OF THE STUDY SPONSOR

This study was coordinated and carried out by site investigators and study coordinators in the participating centers of the 21 Oxybate SXB-26 FMS Study Group and was an industry-sponsored study supported by Orphan Pharmaceuticals, a wholly owned subsidiary of Jazz Pharmaceuticals, Inc. Staff members at Orphan Pharmaceuticals proposed the study but were not involved in recruiting the patients, participating in the statistical analysis, assisting in manuscript preparation, or approving the final proof prior to submission.

Acknowledgements

We wish to acknowledge the developmental work and guidance provided by the research leadership at Orphan Pharmaceuticals (Bill Houghton, MD, PhD, Eric Batson, MD, PhD, Carl S. Hornfeldt, PhD, RPh, and Harry Cook) as well as the editorial and statistical guidance in data analysis and manuscript preparation provided by staff at Jazz Pharmaceuticals, Inc. (Chinglin Lai, PhD, Daniel Pardi, MS, Deborah Waxman, Diane Guinta, PhD, and Susanna Grzeschik, PhD, RPh).

The 21 Oxybate SXB-26 FMS Study Group participating centers, all in the US, and their principle investigators and research coordinators are as follows: Robert M. Bennett, MD, FRCP, and Michelle Price, Oregon Health Science University, Portland, OR; Adam Barron, MD, and Tammy Evans, Radiant Research West Palm Beach, West Palm Beach, FL; Isam Diab, MD, and John Lacombe, Middleburg Heights, OH; Joseph S. Habros, MD, and Heather Yatabe, Radiant Research Scottsdale, Scottsdale, AZ; Andrew J. Holman, MD, and Robin Meyers, Pacific Rheumatology Associates, Inc., Renton, WA; Alan Kivitz, MD, and Deb Morrisson, Altoona Center for Clinical Research, Duncansville, PA; Elliot Kopp, MD, and Jane Downs, Raleigh Neurology, Raleigh, NC; Philip Mease, MD, and Debbie Granner, Seattle Rheumatology Associates, Swedish Rheumatology Research, Seattle, WA; Abigail Neiman, MD, and Dawn Fanguy, Katy, TX; David Nordstrom, MD, and Robert “Bob” Sturgeon, Lynn Institute of the Rockies, Colorado Springs, CO; John Pappas, MD, and Judy Wilkinson, Central Kentucky Research Associates, Lexington, KY; Ashwin Patkar, MD, and Kathryn Tarter, Duke University Medical Center, Durham, NC; I. Jon Russell, MD, PhD, and Wanda Haynes, RN, BSN, CCRC, University of Texas Health Science Center at San Antonio; David Seiden, MD, and Flora Rosen, Broward Research Group, Pembrooke Pines, FL; Eric A. Sheldon, MD, and Mirnaya Alabaci, Miami Research Associates, Miami, FL; Stuart L. Silverman, MD, and Carol Joseph, OMC Research Center, Beverly Hills, CA; N. Lee Smith, MD, and Tami Francisco, Stress Medicine Clinic, Health South Rehab Hospital, Sandy, UT; Daniel Wallace, MD, and Inga Arnold, Wallace Rheumatic Study Center, Los Angeles, CA; Larry G. Willis, MD, and Andrea Craddock, MD, Lynn Health Science Institute, Oklahoma City, OK; John B. Winfield, MD, and Patricia Bradshaw, Alvin Daughtridge Arthritis Center, Lenoir, NC; Patrick Wood, MD, and Laura Warren, Psychopharmacology Research Clinic, Louisiana State University Health Sciences Center School of Medicine, Shreveport.