Table 3.

Summary of anticonvulsants as potential treatment options for diabetic peripheral neuropathy [63–65, 69, 70]

Treatment	Mechanism of action	Advantages	Disadvantages	NNT (95 % CI)	NNH (95 % CI)
Anticonvulsants				3.7 (2.6–6.4)	6.6 (4.9–10.0)
Carbamazepine	Blockage of voltage-gated sodium channels	• Efficacious for pDPN in small studies	• Not FDA-approved for pDPN • Efficacy limited/not evident in large placebo-controlled studies of DSPN • Common side effects include dizziness, ataxia, sedation, hyponatremia, blurred vision, and confusion in the elderly
Lamotrigine	As for carbamazepine	• No clear advantages have been demonstrated for lamotrigine	• Not FDA-approved for pDPN • Limited evidence of efficacy • Common side effects include sedation, dizziness, and ataxia • Rare side effect of Stevens–Johnson syndrome
Pregabalin	Active at the alpha-2-delta subunit of calcium channels and decrease calcium influx	• FDA-approved for pDPN • Efficacious in several neuropathic pain conditions • Dose-dependent effects • Improves QoL scores • Improved tolerability over older anticonvulsants (e.g., carbamazepine) • First-line agent • Metabolized via the kidneys not the liver so reduced potential for drug–drug interactions	• Requires titration schedule • Side effects include somnolence, dizziness, weight gain, headache, dry mouth, and peripheral edema • Dosage reduction necessary in patients with renal insufficiency • Discontinuation rates of 0 %–20 %	4.5 (3.6–5.9)	10.6 (8.7–14)
Gabapentin	As for pregabalin	• Efficacious in several neuropathic pain conditions but effect size is small • Improved tolerability over older anticonvulsants (e.g., carbamazepine) • First-line agent • May provide faster analgesia than pregabalin • Metabolized via the kidneys not the liver so reduced potential for drug–drug interactions • Generics available • Inexpensive	• Not FDA-approved for pDPN • Requires titration schedule • Side effects include dizziness, somnolence, and ataxia • Dosage reduction necessary in patients with renal insufficiency • Adequate trial of treatment can take ≥ 2 months	6.4 (4.3–12)	32.5 (18–222)

CI confidence interval, FDA Food and Drug Administration, NNH the number of patients needed to harm for one drop-out due to adverse events, NNT estimated number of patients with painful polyneuropathy needed to treat to achieve one patient with a 50 % reduction in pain, pDPN painful diabetic peripheral neuropathy, QoL quality of life