Authors: Darwin Luna, Gabrielle Hettie, Luke Pirrotta, Vafi Salmasi, Jennifer M. Hah
Categories: Preliminary Communication, Neuromodulation, analgesia, pain management, peripheral nerve stimulation, neuropathic pain
Source: Pain Management
Authors: Darwin Luna, Gabrielle Hettie, Luke Pirrotta, Vafi Salmasi, Jennifer M. Hah
We aimed to evaluate real-world outcomes of peripheral nerve stimulation (PNS) used to treat chronic neuropathic pain (CNP) at a tertiary pain management center.
Thirty adults who underwent PNS for CNP between June 2015 and September 2021 completed pain and psychosocial assessments in the 6 months before, and 2–3 years after PNS treatment. Pain intensity was measured using the NIH Patient Reported Outcomes Measurement Information System (PROMIS) Pain Intensity Short From (3A). Psychosocial outcomes including depression, anxiety, and sleep disturbance were also measured.
Prior to receiving PNS, long-term responders reported significantly fewer depressive symptoms compared to non-responders (PROMIS depression t-score 50.3 [10.7] vs 57.9 [8.9]; p-value = 0.05). Eleven participants (36.7%) reported long-term treatment response. There was a significantly greater improvement in pain intensity among responders compared to non-responders who reported increased pain (PROMIS Pain Intensity score −9.0 [−4.2] vs. +3.1[+3.2]; p-value < 0.0001).
Patients report clinically meaningful long-term pain relief after receiving PNS through both 60-day and permanent implant systems, with significant reductions in pain intensity observed in long-term responders. Long-term responders reported fewer depressive symptoms compared to non-responders prior to receiving therapy, emphasizing the importance of psychological screening and psychological optimization prior to receiving PNS.
Refractory chronic neuropathic pain treatment is challenging with large variations in efficacy [1]. Chronic neuropathic pain treated with prescription opioids increases the risk of opioid-related harms [2]. For first-line anti-neuropathic medications, the number needed to treat for 50% pain relief ranges from 4 to 10 [2]. Chronic neuropathic pain (CNP) is also associated with a myriad of poor outcomes, including severe disability, poor quality of life, depression, and disordered sleep [3,4].
FDA-cleared percutaneous peripheral nerve stimulation (PNS) systems for the treatment of refractory chronic neuropathic pain are associated with reports of sustained efficacy [5,6]. Previous literature examining off-label use of non-PNS specific leads found 62%–78% of patients with severe, CNP have reported long-term pain relief after 3–16 years of PNS [7,8]. Small off-label studies report near complete pain relief with PNS among patients with CNP after peripheral nerve injury [8–10]. There is a need to characterize long-term outcomes of modern FDA-cleared percutaneous PNS devices for the treatment of CNP.
Much remains unknown regarding predictors of long-term PNS response [11–13]. Traditional frequency (100 hz) stimulation induces comfortable sensations for post-amputation pain [14,15], and post-surgical pain [14,16–18]. Lower frequency (12 hz) stimulation induces comfortable muscle contractions, providing pain relief for subacromial impingement [19], chronic axial low back pain [20], and chronic hemiplegic shoulder pain [21,22].
One RCT reported the efficacy and safety of a percutaneous PNS device (StimRouter, Bioventus) for the treatment of chronic neuropathic pain after peripheral nerve injury [23]. Ninety-four patients were randomized to PNS and conventional medical management vs. conventional medical management alone [6]. A significantly higher percentage of patients receiving PNS (38% vs.10%) were treatment responders at 3 months. Significant improvements in pain interference, quality of life, patient satisfaction, and depressive symptoms, with no serious adverse events, were noted in the PNS group [6]. In a separate multicenter RCT, 28 lower extremity amputees with postamputation pain were randomized to receive femoral and sciatic nerve PNS (SPRINT, SPR Therapeutics) or placebo for 60 days [24]. A significantly higher percentage of patients receiving PNS (58% vs. 14%) reported greater than or equal to 50% pain reduction after 4 weeks of treatment [24]. Sustained relief after PNS for the treatment of chronic low back pain has been reported after 4 months [25]. Of 39 patients receiving the StimRouter system for focal mononeuropathy, a majority reported pain improvement, increased activity, and reduced opioid consumption after 3–6 months [26]. In a retrospective review of 6,160 patients receiving the SPRINTsystem reporting to the device manufacturer’s database, 71% of patients were treatment responders with improved quality of life at the end of the 60-day PNS treatment [27].
Our goal was to examine the real-world long-term efficacy and patient-reported outcomes of percutaneous PNS systems. In addition, we sought to identify patient characteristics associated with long-term treatment response, and to describe the experience of patients receiving PNS for the treatment of chronic neuropathic pain.
This prospective observational cohort study included adult patients with CNP aged 18 years and older who underwent PNS using an FDA-cleared system at a single tertiary academic medical center between June 2015, to September 2021. Additional inclusion criteria were English-speaking participants who had prospectively completed baseline Collaborative Health Outcomes Information Registry (CHOIR) assessments. CHOIR is an open-source learning health care system platform with integrated computerized adaptive testing (CAT) which reduces response burden while yielding greater precision in domain assessment. CAT reduces participant burden through selection of a subset of items from an item bank until measurement meets pre-set criteria for standard errors. CHOIR allows for comprehensive evaluation of biopsychosocial patient-reported outcomes for chronic pain in real-world clinical samples [28]. This study recruited patients who had completed the baseline survey, which included self-reported demographics, clinical characteristics, and psychosocial factors, in the 6 months prior to receiving the PNS implant. Enrolled individuals also agreed to future contact for research studies related to CHOIR at the time of baseline completion. Finally, permission from the patient’s provider at the Stanford Pain Management Center was required to participate in the follow-up survey.
Patients were initially screened for eligibility and consented over the phone by a member of the research team after indicating interest in an e-mail invitation to participate in the follow-up survey. Participants completed a one-time battery of web-based assessments including clinical characteristics and psychosocial assessments. All study procedures were approved by the Institutional Review Board at the Stanford University School of Medicine (IRB #62244).
The present study examined patient self-reported pain intensity, interference, and pain location. This study using the same pain interference and intensity measures for follow-up that were used at baseline prior to PNS therapy. Patients self-reported their pain intensity using the NIH Patient-Reported Outcomes Measurement Information System (PROMIS) Pain Intensity Short Form (3A). Current, worst, and average pain intensity are rated using a five-point Likert scale (0=no pain, 5=very severe) over the past 7 days. The PROMIS Pain Intensity Short form score and other PROMIS measures are reported as a T-score calibrated to the general population in the United States [29]. The minimally clinically important difference in PROMIS pain instruments ranges between 2 to 3 points among patients with chronic pain [30,31]. Thus, long-term treatment responders were characterized as those with ≥2-point decrease in their PROMIS Pain Intensity score from the baseline to follow-up survey. Clinical pain interference in daily activities was assessed with the PROMIS Pain Interference scale [32].
The CHOIR Male or Female Body Map was assessed at the follow-up survey. It consists of an electronic, visual representation of the body that enables patients to indicate locations of pain. There are 36 anterior segments and 38 posterior segments on the CBM [33].
PROMIS measures [34] were used to assess depression and anxiety [35] symptoms, sleep impairment [36], and sleep disturbance [36]. Validated item banks measuring key symptoms allows for interpretable clinical trial results centering on patient-reported outcomes [37–39]. Item banks were administered using the CHOIR-CAT algorithm for the baseline and follow-up surveys. Instruments utilize item response theory yielding standardized T-scores (mean = 50, SD = 10) [40]; and these scores are normed on a large sample of the US population [37].
Patients self-reported age, gender, race/ethnicity, education, and marital status at baseline.
Researchers conducted chart reviews of the electronic medical record (EMR) to determine indications for PNS, PNS device type, PNS implant location, and adverse events not reported by patients in the follow-up survey. The SPRINT Peripheral Nerve Stimulator System (SPR Therapeutics) is intended as a 60-day therapy, after which the percutaneous lead is removed. The StimRouter Peripheral Nerve Stimulator system (Bioventus) utilizes a permanently implanted lead with a conductive electrode that is inserted subcutaneously near the targeted peripheral nerves. An external pulse generator is adhered to the skin, allowing for the device to be powered without a direct connection. Study participants received either the SPRINT or StimRouter PNS System without any crossover. Thus, patients receiving the StimRouter PNS system did not undergo a trial of PNS prior to permanent implant.
All analyses were conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA). Baseline characteristics were summarized with frequencies and percentages for categorical data and means plus standard deviations for continuous data except the number of shaded cells on the CHOIR Body Map, which was reported as median plus interquartile ranges. Differences in baseline demographic and clinical characteristics between PNS long-term responders and non-responders were compared using the t-test for continuous outcomes and the Fisher’s exact test for categorical outcomes. Similarly, differences in follow-up clinical characteristics between PNS long-term responders and non-responders were compared using the t-test for continuous outcomes, and the Fisher’s exact test for categorical outcomes. The Wilcoxon rank sum test was used to assess the difference in the number of shaded cells on the CHOIR Body Map as this variable was not normally distributed. For all comparisons, a P-value <0.05 was utilized for statistical significance.
A total of 30 participants were included in this study. Figure 1 illustrates the study flow for identifying the study cohort. Among them, 36.7% (n = 11) were long-term responders. Among long-term responders, the average change in PROMIS pain intensity scores was a reduction in 9 points, far exceeding the threshold of a 2-point clinically meaningful change. Among long-term responders, 27.3% (n = 3) reported complete pain resolution. Figure 1.Study flow from patients approached to final analysis sample.
The baseline clinical features and demographics of the cohort are summarized in Table 1. Between long-term responders and non-responders, there were no significant differences in age, gender, race, ethnicity, education, marital status, or stimulator manufacturer. The mean age of participants was 57.0 years for responders and 52.2 years for non-responders (p = 0.51). Most participants were white (81.8% of responders and 57.9% of non-responders). Long-term responders reported significantly lower depression symptoms at baseline (p = 0.05). Comparing long-term responders and non-responders, there were no differences in PNS target nerve or indication as summarized in Table 2.Table 1.Baseline clinical features and demographics of the study cohort. Long-Term Responder (n = 11)Long-Term Non-Responder (n = 19)p value^a^Age, mean years (SD)57.0 (12.5)52.2 (21.8)0.51Female gender, n (%)6 (60.0)12 (63.2)1.00Race, n (%) 0.60 White9 (81.8)11 (57.9) American Indian or Alaskan Native0 (0.0)1 (5.3) Asian or Pacific Islander0 (0.0)3 (15.8) African American1 (9.1)0 (0.0) Other1 (9.1)4 (21.1) Hispanic ethnicity, n (%)2 (18.2)2 (10.5)0.61Education, n (%) 1.00 Up to high school diploma or GED0 (0.0)1 (5.3) Up to Bachelor’s degree or Associate degree6 (60.0)11 (57.9) Master’s, Professional, or Doctoral degree4 (40.0)7 (36.8) Marital Status, n (%) 1.00 Never married2 (18.2)4 (22.2) Married or Domestic Partnership8 (72.7)11 (61.1) Divorced, Separated1 (9.1)2 (11.1) Widowed0 (0.0)1 (5.6) Stimulator Company, n (%) 1.00 SPR Therapeutics2 (18.2)5 (26.3) Bioness9 (81.8)14 (73.7) PROMIS T-score Pain Intensity, mean (SD)63.6 (4.5)61.7 (6.2)0.39PROMIS T-score Pain Interference, mean (SD)65.6 (6.7)67.0 (4.9)0.52PROMIS T-score Depression, mean (SD)50.3 (10.7)57.9 (8.9)0.05PROMIS T-score Anxiety, mean (SD)53.2 (11.2)57.5 (9.2)0.26PROMIS T-score Sleep Impairment mean (SD)56.8 (13.7)58.3 (10.4)0.74PROMIS T-score Sleep Disturbance mean (SD)59.8 (9.7)59.4 (7.6)0.90n: Number; SD: Standard Deviation; PROMIS: Patient-Reported Outcomes Measurement Information System.^a^p value for comparisons between responders and non-responders; T-test for continuous comparisons and Fisher’s exact test for categorical comparisons.Table 2.Indications and nerve targets for peripheral nerve stimulation (PNS). Long-Term Responder (n = 11)Long-Term Non-Responder (n = 19)p value^a^PNS Target Nerve, n (%) 0.97 Brachial Plexopathy0 (0.0)2 (10.5) Common peroneal1 (9.1)2 (10.5) Femoral1 (9.1)0 (0.0) Genitofemoral0 (0.0)1 (5.3) Ilioinguinal0 (0.0)1 (5.3) Intercostal0 (0.0)1 (5.3) Lateral femoral cutaneous0 (0.0)1 (5.3) Long thoracic1 (9.1)1 (5.3) Multifidus1 (9.1)1 (5.3) Posterior femoral cutaneous1 (9.1)0 (0.0) Radial1 (9.1)0 (0.0) Saphenous1 (9.1)3 (15.8)* Sciatic1 (9.1)2 (10.5) Suprascapular1 (9.1)0 (0.0) Sural1 (9.1)1 (5.3) Tibial1 (9.1)3 (15.8) PNS Target Location, n (%) 0.94 Back2 (18.2)3 (15.8) Groin1 (9.1)3 (15.8) Lower extremity6 (54.6)11 (57.9) Upper extremity2 (18.2)2 (10.5) PNS Indication 0.19 Brachial Plexopathy0 (0.0)1 (5.3) Complex Regional Pain Syndrome1 (9.1)6 (31.6) Failed back surgical Syndrome1 (9.1)0 (0.0) Foot pain0 (0.0)2 (10.5) Groin pain0 (0.0)1 (5.3) Low back pain0 (0.0)1 (5.3) Neuropathy8 (72.7)6 (31.6) Post-amputation pain0 (0.0)1 (5.3) Shoulder pain1 (9.1)0 (0.0) Thoracic pain0 (0.0)1 (5.3) *1 Patient with both Saphenous and Sciatic Nerve Targets.
Overall, 56.7% (n = 17) of participants received PNS targeting the lower extremities, and 46.7% (n = 14) received PNS for neuropathy.
The clinical features at long-term follow-up are displayed in Table 3.Table 3.Clinical features at long-term follow-up. Long-Term Responder (n = 11)Long-Term Non-Responder (n = 20)p value^a^Days Until Follow-up Assessment, mean (SD)1183.6 (446.5)739.6 (472.2)0.02PROMIS Pain Intensity T-score change, mean (SD)−9.0 (4.2)3.1 (3.2)<0.0001PROMIS T-score Pain Intensity, mean (SD)54.6 (5.8)64.8 (5.9)<0.0001PROMIS T-score Pain Interference, mean (SD)55.3 (5.7)63.6 (7.5)0.004CHOIR Body Map Number of Shaded Cells, median (IQ range)3.0 (1.0–4.0)5.0 (4.0–18.0)0.01^b^PROMIS T-score Depression, mean (SD)44.2 (8.7)50.9 (9.5)0.06PROMIS T-score Anxiety, mean (SD)48.0 (7.1)50.5 (8.6)0.42PROMIS T-score Sleep Impairment mean (SD)50.8 (13.5)57.1 (11.7)0.19PROMIS T-score Sleep Disturbance mean (SD)51.4 (7.5)54.6 (11.9)0.43Quality of life Improved8 (72.7)9 (47.4)0.26PROMIS: Patient-Reported Outcomes Measurement Information System; CHOIR: Collaborative Health Outcomes Information Registry.^a^p value for comparisons between responders and non-responders; T-test for continuous comparisons and Fisher’s exact test for categorical comparisons unless otherwise noted.^b^p value for comparisons between responders and non-responders. Wilcoxon rank sum test for continuous comparisons.
Long-term responders had significantly longer average days until follow-up assessment (1183.6 days) compared to long-term non-responders (739.6 days) (p = 0.02). From baseline to follow-up, there was a significant difference in PROMIS Pain Intensity T-score change in the long-term responder group, with a mean reduction of 9.0 compared to a mean increase in pain of 3.1 in the long-term non-responder group (p = < 0.0001). The median number of shaded cells on the CHOIR Body Map at follow-up for responders was 3.0, while non-responders reported a median of 5.0 shaded cells (p = 0.01) indicating more localized pain among long-term responders to PNS. Figure 2 illustrates reported pain locations on the CHOIR Body Map in the cohort stratified by males and females. It is notable that many female participants reported co-morbid pain in the head region which was not directly targeted by the PNS therapy. Figure 2.Male and female body map visualization. Male body maps (left) and female body maps (right) can be seen in the figure above. Darker shaded regions represent a higher number of respondents who selected the area. Body maps from each gender were layered on top of each other to highlight pain spread in our participant population.
PROMIS T-scores indicated significant improvements from baseline to follow-up in pain intensity and interference for responders when compared to non-responders. Responders had lower pain intensity (54.6 vs 64.8, p < 0.0001) and pain interference T-scores (55.3 vs 63.6, p < 0.0001) at follow-up. Although non-significant, long-term responders reported less depressive symptoms at follow-up compared to non-responders. A 72.7% of long-term responders reported that their quality of life had improved at follow-up compared to 47.4% of non-responders, but the difference was not significant.
For non-responders, most patients reported using the device for several hours a day. Some participants reported using the device for 8–12 h a day. One participant described initial daily use after permanent PNS implantation that was reduced to several days a week as pain improved. Another participant described using the device once a month for 2–3 days at a time. Once the PNS device is turned off 6 participants described continued pain relief lasting from 15 min to 1 week.
For responders, most patients reported using the device continuously. One participant reported using the device once a month. Another participant reported using the device for a few days every 3 to 6 months. Once the PNS device is turned off 5 participants described continued pain relief lasting from a few hours to several weeks. Patient-reported device complications are listed in Table 4, with 1 responder and 3 non-responders noting device malfunction (9.0% and 15.8%, respectively).Table 4.Patient-reported device complications. Long-Term Responder (n = 11)Long-Term Non-Responder (n = 19)Device malfunction1 (9.0)3 (15.8)Retained lead fragment after 60-day PNS0 (0.0)2 (10.5)Adhesive pruritis0 (0.0)1 (5.3)Electric shocks0 (0.0)1 (5.3)
Our questionnaires included forms allowing patients to share their experience with researchers. Some statements “[My device] lowered my pain from a high 8 to a low 3,” and “Prior to having the device I was taking medication to relieve the pain. But I know that within minutes of [the device] starting the pain slowly decreases. I feel relief.” Additionally, participants provided reasons for discontinuing “My pain in my stump has reduced considerably after using [the] device for 3 months after it was inserted,” “I don’t feel the need to use it as I have gotten all of the relief it offered,” and “Initially I used it daily, multiple times during the day, but within the last year, it is down to about 3 days per week.”
Overall, 36.7% of patients receiving PNS in our cohort reported long-term treatment response after 2 to 3 years. These real-world outcomes demonstrate sustained significant and clinically important reductions in pain intensity and pain interference among long-term treatment responders. The longer length of time from PNS therapy to follow-up among treatment responders compared to non-responders in our cohort further highlights the durability of treatment efficacy among responders. Our findings complement prior research demonstrating sustained efficacy of PNS with off-label surgically implanted leads even after several decades [41]. Among patients who had received surgical PNS implant 18–29 years prior to follow-up, periods of active stimulation are associated with reduced spontaneous and heat-induced activity within the pain matrix and increased activation of brain areas involved in pain modulation [42]. The acute effects of PNS on the brain decades after PNS implant signal an underlying mechanism for long-term treatment response and implies a lack of long-term tolerance to the therapy.
At baseline, long-term responders reported significantly fewer depressive symptoms compared to non-responders. The average PROMIS depression t-score was 50.3 vs 57.9 comparing long-term responders to non-responders. Reasonable cut-points for mild and moderate depression are scores of 55 and 60, respectively, when comparing the PROMIS depression T-scores to legacy instruments [43]. These findings suggest the importance of preoperative psychological screening for PNS candidacy. Preoperative psychological interventions to reduce depressive symptoms may enhance PNS treatment response. PNS therapy in and of itself is also associated with improvements in depressive symptoms. In a randomized controlled trial of 89 participants receiving either percutaneous PNS with a micro-implantable pulse generator combined with conventional medical management vs. conventional medical management alone, patients receiving PNS therapy reported significant treatment response at 3 months (84% vs 3%). Only the patients receiving PNS reported significant improvements in pain severity, pain interference, and depressive symptoms at both 3 and 6 months follow-up [44]. In our study, both groups reported less depressive symptoms at follow-up which may have resulted from continued interdisciplinary pain care in our tertiary pain management center. Future work to understand the contribution of preexisting psychological factors to PNS trial and long-term treatment response is needed to inform evidence-based approaches to psychological optimization for PNS therapy.
At follow-up, long-term PNS responders reported a significantly smaller distribution of pain (3 vs. 5 sites) compared to non-responders on the CHOIR body map. Previously, widespread pain on the body map has been categorized for patients with pelvic pain as three to seven additional pain regions outside the pelvis, while intermediate pain is described as one to two additional pain regions outside the pelvis [45]. Similarly, widespread pain on the CHOIR body map has been categorized for patients with abdominal pain as four to eight additional pain regions beyond the abdomen, while intermediate pain is defined as one to three additional pain sites outside the abdomen [46]. Accordingly, patients reporting long-term treatment response to PNS in our cohort reported intermediate pain, while non-responders reported widespread pain. As the diffuseness of pain is thought to be proportional to the degree of systemic pathophysiology characterized by centralized pain symptoms (e.g., reduced pressure pain thresholds, increased central sensitization), it is possible that long-term PNS responders experienced clinically significant normalization of central sensitization as a result of treatment. In addition, prior research demonstrates normalization of impaired conditioned pain modulation in response to PNS therapy. Conditioned pain modulation is a measure of the efficacy of descending pain pathways in activating endogenous analgesia. Occipital nerve stimulation for intractable headache results in sustained improvements in impaired conditioned pain modulation from 2 weeks to 12 months after PNS implant [47]. However, we did not assess pain distribution at baseline with the CHOIR body map, and it is also possible that patients with less central sensitization and intact conditioned pain modulation at baseline are more likely to respond to PNS. Future research tracking the distribution of pain in response to PNS treatment is warranted.
The study has several limitations, including the small sample size of 30 participants. In addition to a small sample size, this study did not recruit a diverse group of participants. Further research is needed to explore the effectiveness of PNS therapy in racial and ethnic minorities, individuals with disabilities, and patients with lower socioeconomic status. The majority (77%) of our cohort received a StimRouter device, which could introduce bias in our findings as this is a permanently implanted device. However, no difference in treatment response was noted between device groups in our cohort. Future research inclusive of all modern percutaneous PNS systems, with analyses stratified by device type, will allow for more granular characterization of treatment responders and factors associated with PNS treatment efficacy. Similarly, standardization of time to follow-up for assessment of patient-reported outcomes will decrease concerns for non-responder bias. As all patients were receiving care in a tertiary pain management center, outcomes may have been influenced by other pain therapies patients received in during the time to follow-up assessment.
We report real-world long-term outcomes of modern, percutaneous PNS systems for a diverse array of indications and target nerves. Overall, long-term responders did not report elevated depressive symptoms prior to receiving PNS therapy, while non-responders reported symptoms consistent with mild depression. These findings highlight the importance of preoperative psychological screening and optimization for PNS therapy. Patients reported sustained clinical efficacy several years after receiving PNS therapy with reductions in pain intensity, pain interference, and the distribution of pain, and future research is needed to track these long-term outcomes in larger cohorts of patients receiving PNS therapy.