OBJECTIVE: Neuropathic pain is difficult to manage and pharmacotherapy is often unsatisfactory. Patients often will have tried a large number of drugs in an effort to find relief. This study evaluated the effectiveness and safety of topiramate in patients with peripheral diabetic neuropathy (PDN) who have previously tried other anticonvulsants. METHODS: Retrospective review of 100 patient charts using physician interview. Physicians were asked to include the 4 most recent patients that met the following criteria: 1) diagnosis of PDN; 2) topiramate started > 6 months prior to review date to allow an adequate trial time for drug therapy; 3) no concomitant gabapentin use. A subgroup of 28 patients that had failed other anticonvulsants prior to topiramate was analyzed. RESULTS: Patient characteristics: 14% Type 1 diabetes, 86% Type 2; mean age 64.9 years; 54% males; 93% Caucasian. Most commonly affected areas: foot/toe (82%), calf (36%), finger/hand (32%), thigh (21%). Most common qualities: burning (57%), tingling (50%), aching (46%), pins and needles (21%). Previous anticonvulsants included gabapentin (82%), carbamazepine (29%), clonazepam (4%), and divalproex (4%). The most common reason for anticonvulsant discontinuation was lack of efficacy. Patients had PDN for 34 months on average prior to starting topiramate. Mean duration of topiramate therapy was 17 months; mean total daily dose was 127mg. A total of 57% used topiramate monotherapy. According to physician assessment, 46% (95%CI 28, 64) were "very much improved" or "much improved" for pain, 36% (95%CI 18, 54) for physical activity, and 39% (95%CI 21, 57) for sleep. None had worsening of symptoms. A total of 32% experienced topiramate-related adverse effects (AEs). In 67% no action was taken for AEs. There were no discontinuations due to AEs or lack of efficacy. CONCLUSIONS: Topiramate was effective and well-tolerated for the treatment of PDN, even in a group of difficult to treat patients for whom other anticonvulsants had failed.