Absence of long-term analgesic effect from a short-term S-ketamine infusion on fibromyalgia pain: A randomized, prospective, double blind, active placebo-controlled trial

Fibromyalgia is a chronic widespread musculoskeletal pain condition without a clear structural or inflammatory cause, affecting about 2–3% of the population with a strong female predominance. Although its aetiology remains contested, earlier research has associated fibromyalgia with central sensitisation and altered pain processing; N-methyl-D-aspartate receptor (NMDAR) activation is one mechanism implicated in that sensitisation. Small clinical studies and mechanistic work have suggested that blockade of the NMDAR (for example with ketamine or dextromethorphan) can produce short-term analgesia and that an intravenous ketamine response may predict benefit from subsequent oral NMDAR antagonists. This trial set out to test whether a single, short-duration infusion of the S(+) enantiomer of ketamine (S-ketamine) produced pain relief in fibromyalgia that outlasted the pharmacokinetic presence of the drug. Specifically, the investigators compared a 30-minute infusion of 0.5 mg/kg S-ketamine to an active placebo (5 mg midazolam) and hypothesised that ketamine would (i) provide greater pain relief than placebo and (ii) produce analgesia that persisted beyond the treatment period. An active placebo was chosen to help maintain blinding by producing sedation and some acute subjective effects during the infusion.

The study was a randomised, prospective, double-blind, active placebo-controlled trial conducted after ethical approval and registered prior to recruitment. Fibromyalgia patients were recruited from outpatient clinics and a patient association website. Inclusion required a diagnosis based on the 1990 American College of Rheumatology criteria, age 18–75 years and a baseline spontaneous pain score of 5 or greater on a 0–10 visual analogue scale (VAS). Key exclusions included body weight > 100 kg or BMI ≥ 35, use of strong opioids, substantial anxiety/depression (HADS subscale ≥ 8), pregnancy, severe cardiovascular or psychiatric disease, and other comorbidities incompatible with study drugs. Patients could continue stable background therapies during the 8-week study but not change them. An independent physician randomised participants and prepared blinded syringes. On a single treatment day patients received either 0.5 mg/kg S-ketamine or 5 mg midazolam given over 30 minutes, followed by 2.5 hours of acute measurements and weekly follow-up for 8 weeks. Two intravenous lines were placed for drug delivery and blood sampling. Those allocated to midazolam were offered an optional open-label ketamine treatment after study unblinding. Primary and secondary assessments included spontaneous fibromyalgia pain measured by VAS (three baseline measures averaged, then at 45, 60, 75, 90, 120, 150 and 180 minutes after infusion start), an experimental heat pain test using a thermode with individually titrated peak temperature (VAS for heat pain at baseline, 45, 60, 120 and 180 minutes), and side effects assessed with the Bowdle questionnaire (13 psychotomimetic items scored 0–10 at baseline, 45, 60, 120 and 180 minutes). Vital signs were monitored throughout. Venous blood sampling for plasma S-ketamine and S-norketamine concentrations was performed frequently during the 3-hour period and analysed by HPLC. For longer-term impact the Fibromyalgia Impact Questionnaire (FIQ) was completed weekly for 8 weeks to capture pain, stiffness and functional status over the preceding week. The study was powered to detect a 2-point difference in pain score during follow-up, assuming SD 1.5, power 0.8 and α = 0.05, yielding a planned sample of 10 per group increased to 12 per group to allow attrition. Analysis followed the intention-to-treat principle (that is, all randomised patients were analysed in their assigned groups). Demographics and baseline comparisons used t-tests or Chi-square/Fisher exact tests as appropriate. A responder was defined as > 50% reduction in spontaneous pain. A linear mixed model was used to analyse repeated measures (pain scores, side effects and heat pain), with time as a within-subject factor and treatment as a between-subject factor. Change scores relative to baseline (ΔVAS) were specifically compared at 45 and 180 minutes. P-values < 0.05 were considered statistically significant.

Recruitment flow: 63 patients were approached, 27 declined, 36 were screened and 25 were randomised; one patient withdrew before baseline, leaving 24 treated participants (12 per group). Median time since diagnosis was 1.3 years (range 0.1–16 years) and the groups had on average 16 tender points; baseline demographics and values were reported in a table (not reproduced in the extracted text). Spontaneous fibromyalgia pain: Baseline VAS scores were 5.4 ± 0.6 cm (S-ketamine) and 5.8 ± 0.4 cm (midazolam), not significantly different. After infusion S-ketamine reduced pain to 1.9 ± 0.8 cm at 45 minutes (P < 0.01 versus baseline) and 3.1 ± 0.8 cm at 180 minutes (P < 0.01 versus baseline). Midazolam pain scores were 4.4 ± 0.6 cm at 45 minutes (P < 0.01 versus baseline) and 4.3 ± 0.7 cm at 180 minutes (P < 0.01 versus baseline). A significant main effect of time was observed (P < 0.001) but there were no statistically significant main effects of treatment (P = 0.09) or time-by-treatment interaction (P = 0.10), indicating no clear difference between S-ketamine and midazolam across the 45–180 minute recovery period. VAS trajectories in the S-ketamine group tracked measured plasma S-ketamine concentrations, consistent with a pharmacokinetic-driven effect. Responder analysis showed more early responders with S-ketamine: at 45 minutes eight patients in the S-ketamine arm versus three in the midazolam arm had > 50% pain reduction. Among responders at 45 minutes, S-ketamine reduced mean pain from 5.3 ± 0.8 to 0.6 ± 0.2 cm, while midazolam responders reduced from 5.2 ± 1.9 to 1.8 ± 1.0 cm. The number of responders declined over time: at 180 minutes there were six responders in the S-ketamine group (mean pain 0.8 ± 0.3 cm) and three in the midazolam group (2.0 ± 0.3 cm). Experimental heat pain: Baseline heat pain VAS differed between groups (6.5 ± 0.3 cm S-ketamine vs 5.4 ± 0.4 cm midazolam, P = 0.046), so results were analysed relative to baseline. S-ketamine produced a mean reduction of 2.0 ± 0.9 cm at 45 minutes (P < 0.01 versus baseline) and 0.2 ± 0.7 cm at 180 minutes (not significant). Midazolam effects were −0.9 ± 0.7 cm at 45 minutes (P < 0.05) and +0.5 ± 0.6 cm at 180 minutes (not significant). There was a significant main effect of time (P < 0.01) but no significant group or interaction effects, indicating no reliable difference between treatments on experimental heat pain in the 2.5 hours post-infusion. ΔVAS comparisons showed no significant differences between experimental and fibromyalgia pain at 45 or 180 minutes for either treatment. Side effects and monitoring: Pre-treatment Bowdle scores were similar between groups. After infusion five of the 13 Bowdle items (surroundings, time, reality, thoughts and high) showed significantly greater effects with S-ketamine at 45 minutes (P < 0.05); the largest difference was for the reality item (6.5 ± 1.0 with S-ketamine versus 1.1 ± 0.7 with midazolam at 45 minutes). The total Bowdle scale did not show a main treatment effect (P = 0.08), but time (P < 0.01) and time-by-treatment (P < 0.01) effects were present. Vital signs remained within clinically acceptable ranges; during S-ketamine infusion heart rate and blood pressure increased by approximately 20% and respiratory rate decreased by about 10%, whereas oxygen saturation stayed > 98% in both groups. Longer-term outcomes (FIQ weeks 1–8): Baseline FIQ scores were similar (52 ± 4 S-ketamine; 50 ± 3 midazolam). No significant time (P = 0.07), group (P = 0.98) or interaction (P = 0.80) effects were observed over the 8-week follow-up. FIQ-derived pain scores likewise showed no time, group or interaction effects (P = 0.09, 0.55 and 0.57 respectively) and the number of responders by FIQ pain averaged two patients per group. Open-label S-ketamine given to midazolam patients after unblinding did not show different FIQ outcomes compared with the RCT data (details not shown).

Goldenberg and colleagues conclude that a single 30-minute infusion of 0.5 mg/kg S-ketamine produced no significant analgesic benefit over an active midazolam control on spontaneous or experimental pain in fibromyalgia patients during the hours after treatment or over an 8-week follow-up. They note a transient signal in the first 45 minutes—more early responders in the S-ketamine group—but emphasise that overall effects closely followed plasma ketamine concentrations, consistent with a pharmacokinetic rather than a lasting pharmacodynamic effect. When placed alongside previous randomised trials that used shorter or lower-dose ketamine infusions, the present findings point to a limited and largely transient analgesic effect of single short-term ketamine infusions in fibromyalgia. Earlier studies reported substantial pain reductions during infusion and, in some cases, brief post-infusion benefit lasting days, but none demonstrate consistent long-term relief after a single short infusion. The authors therefore suggest that infusion duration rather than peak dose may be the critical determinant for producing longer-lasting analgesia; longer continuous infusions in other neuropathic pain syndromes have produced more prolonged benefit in randomised studies. The investigators discuss alternative explanations: fibromyalgia is likely pathophysiologically heterogeneous, so only a subgroup with NMDAR-mediated sensitisation may respond; imaging work has suggested differing brain perfusion patterns in responders versus non-responders. Methodological considerations are also acknowledged. The use of an active placebo (midazolam) helped maintain blinding by producing sedation, but may have reduced the contrast in analgesic outcomes. Severe sedation during infusion prevented on-treatment pain assessments prior to 45 minutes, so early differences during the infusion could have been missed. A post hoc calculation indicated that the study was underpowered to detect differences in the immediate post-infusion period—approximately 30 patients per group would have been required to show a significant effect in the first three hours. Finally, the authors reiterate that while they reject the hypothesis that a single short high-dose S-ketamine infusion yields long-term analgesia in fibromyalgia, more prolonged or repeated intravenous regimens or oral NMDAR antagonist treatments might still be effective and warrant further study.