Elan: ziconotide review focused on off-label uses.

A. Narayana
DOI: https://doi.org/10.1177/104990910502200602
2005-11-01
Abstract:We read with interest Dr. Eric Prommer’s review of ziconotide in the September/October 2005 issue.1 The majority of references and discussion for this review focused on off-label uses and off-label titration schemes for ziconotide therapy.2,3 The purpose of this letter is to highlight the recommended titration schedule by summarizing the key findings from a recent study.4 During the American Association of Pain Management (AAPM) meeting in 2005, our most recent pivotal study utilizing a slow titration regimen (Study 301) was presented by Dr. Lynn Webster. Study 301 forms the basis of the dosage and administration section of the US package insert. The slow-titration study randomized 220 patients to ziconotide or placebo. This study examined patients who had failed all other pain-therapy options. The mean percentage change in Visual Analog Scale of Pain Intensity (VASPI) score from baseline to end of the initial titration was significantly higher (p = 0.036) for patients in the ziconotide group (14.7 percent) than in the placebo group (7.2 percent). Secondary outcome measures supported the clinical significance of the primary efficacy outcome measure. Statistically significant differences on the Clinical Global Impression (CGI) scale were found: 28.4 percent of patients in the ziconotide group compared with 12.1 percent of patients in the placebo group reported “a lot” or “complete satisfaction” with therapy. Ziconotide therapy was well-tolerated in this study. The discontinuation rate in Study 301 due to adverse events (4.5 percent for ziconotide patients and 4.6 percent for placebo patients) as well as the incidence of serious adverse events (11.6 percent and 9.3 percent, respectively) were not significantly different between ziconotide and placebo. The 11.6 percent serious adverse event rate in the ziconotide arm at three weeks in Study 301 compares with a 30.6 percent serious adverse event rate at the end of initial titration (5 to 6 days) in the ziconotide arm of the Staats study.3 Thus, after taking into consideration the different patient exposures, the incidence of serious adverse events was more than three times greater in the Staats study than that in the slow-titration study. The results of Study 301 demonstrate that there is a much-improved tolerability and safety profile with a slower administration of ziconotide therapy. Over 1,200 patients have been treated with ziconotide in the clinical studies leading to approval.5 First, we have seen no evidence of addiction to ziconotide therapy. Next, no evidence of tolerance development has been seen with ziconotide therapy in preclinical6 and clinical7,8 studies. Nor has evidence of physical dependence or withdrawal been seen with discontinuation of ziconotide.5 In clinical trials, ziconotide did not suppress respiratory function. In contrast, opiates are known to cause lifethreatening respiratory depression. Overdoses of ziconotide do not appear to be lethal. In contrast, overdosage of morphine can be lethal. There has been no evidence of granuloma formation in any of our ziconotide studies. In contrast, high concentrations of morphine can cause granulomas leading to irreversible neurologic deficits, including paralysis.9 In summary, intrathecal ziconotide is the first treatment for patients with severe chronic pain in 20 years supported by double-blind, placebo-controlled clinical studies. I encourage your readers to contact me if they have any questions. Sincerely,
What problem does this paper attempt to address?