BackgroundPainful diabetic neuropathy (PDN) is a serious, polymorphic, and prevalent complication of diabetes mellitus. Most PDN treatment guidelines recommend a selection of drugs based on patient comorbidities. Despite the large numbers of medications available, most randomized clinical trials (RCTs) conducted so far have yielded unsatisfactory outcomes. Therefore, treatment may require a personalized approach based on pain phenotype or comorbidities.MethodsTo evaluate whether or not a patient’s pain phenotype or comorbidities can influence the response to a specific PDN treatment, we conducted a systematic review using two different approaches: pain phenotype and associated comorbidities-based treatment.ResultsOut of 45 identified papers, 7 were thoroughly reviewed. We found four RCTs stratified according to pain phenotype with three main results: (1) paroxysmal pain had a better response to pregabalin; (2) the preservation of thermal sensation or nociception anticipated a positive response to the topical treatment of pain; and, (3) after a failure to duloxetine (60 mg/day), the patients with evoked pain or severe deep pain had a better response to association of duloxetine/pregabalin while those with paresthesia/dysesthesia benefited from duloxetine monotherapy (120 mg/day). By contrast, the other three papers provided weak and even contradictory evidence about PDN treatment based on comorbidities.ConclusionAlthough more studies are needed to provide an adequate recommendation for clinical practice, our systematic review has provided some evidence that PDN phenotyping may optimize clinical outcomes and could, in the future, lead to both less empirical medicine and more personalized pain therapeutics.