Deep brain stimulation of the subthalamic nucleus: taking the ouch out of Parkinson disease.

Abstract

Parkinsondisease (PD) is knowntoaffectmultipleneurotransmitter systems, resulting in both motor and nonmotor symptoms. Pain is an underrecognized but important nonmotor symptom of PD that may be verydisablingandisknownto impair quality of life.1 While significantprogresshasbeenmadeinamelioratingmotorsymptomsusing pharmacological, rehabilitative, and surgical treatments, thetreatmentofpain inthiscontexthasnotreceivedsufficientattention.Aliteraturereview2notedareportedprevalence ofpain inPDof40%to85%,yetonlyabouthalf ofpatientswith PDwhofeltpaintookanalgesics.ClassifyingPD-relatedpaininto its subtypes3 is important for rational treatment, and themost common subtypes of pain in PD are musculoskeletal and dystonic, with central neuropathic pain being the least common.2 Toaddresspainsuccessfully inPD,anunderstandingof itspathogenesis is important, yet this has remained elusive, in part because its causes areprotean. Just as certainmotor symptomsof PD aremore responsive to dopaminergic drugs and deep brain stimulation (DBS), somepainsubtypesmayrespondbetter than others to pharmacological and surgical intervention. Dysfunction of both dopaminergic and nondopaminergic basal ganglia pathwaysare likelytobe involvedinPD-relatedpain,whichmay explainwhysome typesofpain are responsive to levodopaand others are not.4 Deep brain stimulation of the subthalamic nucleus (STN) isnowanestablished treatmentofdisablingmotor symptoms in advanced PD, and it is therefore important to understand its effect on pain associated with PD. Previous reports5,6 on the effects of STNDBSonpain inPDshowthat the levelsofmusculoskeletal anddystonicpaindecreasedwhenassessed1yearafter surgery,but, toourknowledge, there isnoexisting information on the status of chronic pain. In this issue of JAMA Neurology, Jung and colleagues7 investigate the long-termeffect of STNDBSonpain in24patients withPDwhowereobservedfor8yearsafterundergoingtheprocedure for alleviating disabling motor symptoms. The beneficial effectsofSTNDBS in this cohort 3monthsafter surgeryand 24monthsafter surgeryhavebeenpreviouslypublished.8,9The subtypes of painwere classified according to Ford.3 The severity and body distribution of pain were compared between presurgical andpostsurgical states. Jungandcolleagues7notedsignificant improvement at 8 years in all pain types present preoperatively, regardlessofwhether thepreoperativepainwas responsive to levodopa or not.However, newpain, not present before surgery, developed in 75% of patients during prolonged follow-up. Musculoskeletal pain accounted for the majority of newpain cases. Although themean severity of painwas less at 8 years than at baseline,more patients experienced pain at the end of the observation period than preoperatively. Because previous studies on pain following STN DBS for PD are of short duration, the durability of the procedure’s effect on pain is not well established. The chief strength of the work by Jung and colleagues7 is the long follow-up period, which suggests that, althoughDBSmay relievepain for a time, this is not a durable effect owing to the onset of new, primarilymusculoskeletal pain.A recent report5wasconcordantwith the study by Jung and colleagues7 in showing the benefit of DBS on dystonic and musculoskeletal pain but was discordant in showingnoeffect ofDBSoncentral and radicular pain. Methodological differences, the subjective classification of pain, andamuchshorter observationperiod couldaccount for these discrepant findings. The study by Jung and colleagues7 has several limitations that the authors acknowledge, including a small cohort without a control group thatwas treatedmedically, the lack of correlation of pain scores with motor Unified Parkinson’s Disease Rating Scale scores (especially rigidity, which can potentially impact musculoskeletal pain), the lack of measures of disability or effect of pain on quality of life, and the lackofmoodandcognitiveassessments that can influencepain perception. In addition, an analysis of the effect of levodopa onpreoperative pain vs the effect of DBS onpainwas not performed. This would have been of interest in light of a previous report10 suggesting that the effects of DBS on pain can be predicted by measuring the effects of levodopa on pain. Despiteits limitations,thestudybyJungandcolleagues7providesanovelperspectiveon thedurabilityof thepain-relieving properties of STN DBS in PD. The authors direct our attention tothefactthatmusculoskeletalpainmayemergeyearsafterDBS, warranting individualized treatment. The next step is to pursueadeeperunderstandingof themechanismofpain inPD.Previous work11 with small numbers of participants has demonstrated altered pain processing in participants with PD who experienced pain comparedwith those who did not, such that STN DBS raises pain thresholds selectively in those who experiencepain.Althoughthere isagrowingconsensusthatSTNDBS decreases the level of pain in people with PD, the literature is mixed on the subtypes of pain that are responsive to DBS, and the study by Jung and colleagues7 shows that new pain arising years after the procedure is common. This underscores the importance of performing future trialswith larger cohorts, longer observational periods, and standard methods to enable effective interpretation of outcomes. For now,we have learned that STNDBS does not take the ouch out of PD in the long run. Related article page 504 Opinion