The Anatomy of a Successful Trial

Orientation

This reading prepares you for our introductory lecture on spinal cord stimulation. SCS is the most common functional neurosurgery procedure in the United States by an order of magnitude, and the one you are most likely to begin doing independently earliest in your career. It is also the procedure most likely to disappoint you — or your patient — if you get the decision-making wrong. The technical part of putting leads in the epidural space is the easiest part of the whole enterprise.

The central argument of this reading — and the one I will return to repeatedly in lecture — is that patient selection accounts for the large majority of outcome variance in SCS. Waveform choice, device brand, lead trajectory, and surgical finesse matter, but they are fine-tuning around a first decision that, if wrong, nothing downstream can rescue. The best implanters in the country have explant rates in the 15–25% range at five years, and the single biggest driver of those explants is not lead failure — it is patient selection error. Know this before you place your first lead.

Three facts, interlocking, are worth memorizing before anything else in SCS.

First, the mechanism of SCS is incompletely understood. Fifty years after Shealy implanted the first device, we still cannot tell you with certainty why dorsal column stimulation reduces pain in some patients and not others. The gate-control theory that motivated the first implants has been substantially revised; newer models invoke supraspinal processing (insula, ACC, descending modulation), glial mechanisms, and — for ECAP-controlled systems — dose-dependent activation of A-beta fibers. This matters clinically because it means we cannot predict responders from a biomarker. We predict responders from phenotype, expectations, and comorbidities.

Second, the therapy is neuropathic-preferential and sensitive to psychosocial factors in a way most surgical therapies are not. A perfectly placed lead in a patient with somatization, untreated depression, active litigation, or opioid use disorder will fail. A mediocrely placed lead in a well-selected patient will often still work, because the patient’s brain is ready to respond.

Third, the failure mode is visible. Unlike a spinal fusion where the patient quietly continues to have pain and the surgeon rarely hears about it, a failed SCS patient comes back every three months, asks for reprogramming, ends up requesting explant, and often ends up on more opioids than before implant. You will live with your selection decisions. Choose carefully.

Not all SCS indications are created equal. The evidence base is deepest for a handful of conditions and thin-to-absent for many others. Know where you are on the evidence ladder for every patient you implant, and disclose that level of evidence explicitly in the clinic.

Level 1A evidence (multiple RCTs)

Persistent spinal pain syndrome type 2 (PSPS-2, formerly "failed back surgery syndrome"). The foundational indication. Kumar PROCESS (2007) established SCS superiority over conventional medical management; North (2005) showed SCS superior to reoperation. Kapural SENZA-RCT (2015, 2016) established 10 kHz superiority over tonic for back-and-leg pain. Deer SUNBURST (2018) established burst superiority over tonic. Mekhail Evoke (2020, 2022) established ECAP-controlled closed-loop superiority over open-loop. Every newer waveform defines itself in reference to this indication.

Painful diabetic neuropathy (PDN). Petersen SENZA-PDN (2021 JAMA Neurol) demonstrated 10 kHz superiority over conventional medical management in 216 randomized patients refractory to gabapentinoids/duloxetine — 76% responder rate at 6 months vs 5% in CMM arm. This is the indication that changed the conversation for SCS in non-surgical populations and earned FDA approval in 2021. One of the most rapidly growing indications.

Complex regional pain syndrome (CRPS) types I and II. Kemler (2000) established paresthesia-based SCS superiority over physical therapy in CRPS-I; Deer ACCURATE (2017) demonstrated dorsal root ganglion stimulation superior to SCS for CRPS of the lower extremity — the indication where DRG-S has an evidence advantage over dorsal column SCS. Think DRG first for focal CRPS-I below the waist.

Level 1B–2A evidence (single RCT or strong observational)

Non-surgical refractory back pain. Abbott DISTINCT trial (2024) demonstrated burst SCS superior to CMM in non-surgical low back pain. FDA-expanded indication. Useful for the patient who is not a fusion candidate but has imaging-concordant, activity-limiting axial back pain.

Refractory angina pectoris. European registry data and small RCTs support SCS for medication-refractory angina; mechanism putatively anti-ischemic via sympathetic modulation. Uncommon in US practice — cardiology rarely refers. Mention in lecture for completeness.

Post-herpetic neuralgia, post-amputation (phantom limb) pain, peripheral neuropathic pain (entrapment residua, post-traumatic). Case series and small prospective studies support a trial when conservative and interventional management have failed.

Low-quality or absent evidence — handle with skepticism

Cancer pain, central post-stroke pain, brachial plexus avulsion, spinal cord injury pain, MS pain. Case reports and series only. Can be considered in highly selected patients after multidisciplinary discussion, but frame it to the patient as genuinely exploratory.

Fibromyalgia, widespread non-specific pain, somatoform disorders. Do not implant. The evidence is not there, the failure rate is punishing, and the patients are often the ones most injured by the experience of another failed intervention.

This is the section to read twice. Every bullet below represents a patient someone has implanted against better judgment and regretted. Build the habit now of screening rigorously; it will save you and your patients a lot of grief.

Psychological — the single biggest predictor of failure

Every SCS candidate gets a formal psychological evaluation before trial. This is not a checkbox; it is the single most important piece of the preoperative workup. NACC and nearly every payer require it. Use a clinical psychologist experienced with chronic pain and neurostimulation, not a generic pre-surgical mental health screen.

Absolute or near-absolute contraindications: active psychosis; active suicidal ideation; active substance use disorder (including alcohol, opioids, stimulants); untreated severe major depression; active somatization or conversion disorder; malingering; personality disorder with active impulsivity or splitting that has disrupted prior medical care.

Relative contraindications requiring treatment and re-evaluation: moderate depression or anxiety on suboptimal treatment, PTSD without active trauma-focused therapy, recent major psychosocial stressor (divorce, bereavement, job loss within 6 months), pain catastrophizing scores in the severe range without psychotherapy engagement.

Medical Optimization

Smoking cessation. NACC recommends ≥4 weeks abstinence preoperatively. Verify with cotinine testing for high-risk patients. Smoking dramatically increases infection, wound-healing, and pseudarthrosis risk for paddle implants, and correlates with worse pain outcomes independent of surgical complications.

Glycemic control. Target HbA1c <7.5% before elective implant; some centers require <8.0%. Higher values approximately double infection rates in most series. Obvious exception: PDN candidates often have suboptimal control partly because of the pain; in those patients, optimize what you can and proceed when the team agrees risk is acceptable.

BMI and skin integrity. BMI >40 increases wound complications, IPG pocket failures, and lead migration. Do not hard-decline on BMI alone, but have a long pre-op conversation and prefer a flank pocket over upper buttock in obese patients to avoid sitting-surface dehiscence.

Anticoagulation. Follow ASRA interventional pain guidelines and the NACC bleeding/coagulation consensus. Epidural hematoma is the feared complication. Hold anticoagulants at appropriate intervals (warfarin INR <1.4, DOACs 2–5 days depending on agent and renal function, clopidogrel 5–7 days, aspirin generally continued for cardiovascular secondary prevention). Do not restart until wound is dry, typically 24 h postop.

Opioid use. High-dose chronic opioid use (>90–120 MME/day) is a relative contraindication and a well-documented negative predictor of SCS response. This does not mean "refuse to implant anyone on opioids"; it means have an honest conversation about dose reduction as a parallel goal, engage pain psychology, and reconsider implant timing. Post-SCS opioid reduction is a real outcome but not a guaranteed one.

Pain classification. Neuropathic pain responds; pure nociceptive pain responds poorly. Mixed pain (radicular + axial) is the sweet spot, which is why PSPS-2 has been the flagship indication. Screen with DN4 or PainDETECT, document the neuropathic features, and avoid implanting patients whose dominant complaint is mechanical axial pain without radicular or neuropathic features.

The SCS trial is practically universal in US practice — typically 5–7 days of percutaneous lead placement with an external pulse generator, evaluated against a threshold of ≥50% pain reduction plus functional improvement and no increase in opioids. It remains the NACC-endorsed gate to permanent implant. But the evidence for trial necessity has been questioned.

The TRIAL-STIM study (Eldabe 2020) — a UK multicenter RCT of 105 patients — randomized SCS candidates to a screening trial vs no trial before permanent implant and found no meaningful difference in 6-month pain outcomes or cost-effectiveness between the two strategies. More recent NACC guidance preserves the trial as standard-of-care in the US (especially under Medicare coverage rules), but acknowledges that in selected patients — particularly well-studied indications like PDN — the trial may be adding cost without meaningfully improving selection. Expect this to evolve over the next five years.

What the trial is useful for:

• Confirming that the patient’s phenotype responds to stimulation at all.
• Letting the patient experience the therapy, set expectations, and decide whether they want to live with a device.
• Testing paresthesia coverage (tonic waveforms) or confirming analgesic response (sub-perception waveforms).
• Identifying the patients who will explain during the trial that their spouse, dog, or sleep is the actual problem.

What the trial is not useful for:

• Predicting long-term response with high accuracy. Trial-to-permanent agreement is good but imperfect, and short-trial responders still explant at measurable rates.
• Differentiating waveforms in a rigorous way — the trial is generally run on the waveform the patient will receive permanently, not as a head-to-head.

The last decade has seen an explosion of stimulation paradigms. Residents often arrive at lecture believing one waveform is clearly superior. The honest answer is that all modern waveforms outperform conventional tonic stimulation in at least one high-quality trial, and that head-to-head comparisons between novel waveforms are thin. Commercial competition substantially exceeds clinical differentiation. That said, there are sensible defaults.

The Waveform Landscape

Conventional tonic (paresthesia-based), 30–80 Hz, 100–500 μs, 2–6 mA. The legacy waveform. Works by producing paresthesias that mask pain. Intraoperative paresthesia mapping is required. Largely superseded for back-and-leg pain by newer waveforms but still used in cervical, peripheral nerve, and some CRPS cases.

10 kHz high-frequency (Nevro Senza). Paresthesia-free. SENZA-RCT established superiority over tonic for back-and-leg pain at 3 and 24 months; SENZA-PDN demonstrated robust effect in painful diabetic neuropathy. Anatomic placement at T9–T10 without paresthesia mapping — a meaningful workflow advantage for asleep placement. Downside: higher energy consumption, charging burden.

Burst (Abbott, originally DeRidder). Intermittent packets of five high-frequency pulses at 500 Hz, delivered at 40 Hz. Putative dual-mechanism action on medial (affective) and lateral (sensory-discriminative) pain pathways. SUNBURST demonstrated superiority over tonic in a crossover trial — 70% of patients preferred burst. Paresthesia-free at therapeutic settings. Abbott’s proprietary "DeRidder burst" remains the most-studied formulation.

DTM (Differential Target Multiplexed, Medtronic). Multiplexed delivery of two programs simultaneously, designed to target both neurons and glial cells per preclinical work. Superior to conventional SCS in the Medtronic-sponsored RCT for back-and-leg pain. Available on Intellis and Vanta platforms.

ECAP-controlled closed-loop (Saluda Evoke). The most mechanistically interesting modern system. Measures evoked compound action potentials from the dorsal columns in real time and adjusts stimulation intensity pulse-by-pulse to maintain a target neural response. Evoke RCT demonstrated superiority over fixed-output open-loop stimulation at 3, 12, 24, and 36 months. Elegant, but requires a specific platform and is the only option that delivers pulse-by-pulse physiologic feedback.

Brand Orientation (Non-Exhaustive)

• Medtronic (Intellis, Vanta, Inceptiv). Oldest commercial platform. Multiple waveforms including DTM. Inceptiv adds ECAP sensing. Vanta is recharge-free.
• Boston Scientific (WaveWriter Alpha). Multi-waveform (conventional, burst variants, sub-perception 1–10 kHz), 32-contact architecture with 16-contact dual-lead capability, CoverEdge surgical paddle. Strong programming flexibility.
• Abbott (Proclaim XR, Eterna). BurstDR waveform (DeRidder original). Proclaim XR is the primary recharge-free platform. Remote programming via iOS app. FDA-expanded to non-surgical back pain with the DISTINCT trial.
• Nevro (Senza, HFX). 10 kHz high-frequency leader. Anatomic placement workflow. HFX iQ adds AI-driven closed-loop programming. Focused on PDN and PSPS-2.
• Saluda (Evoke). ECAP-controlled closed-loop. Only commercial system that delivers pulse-by-pulse physiologic feedback. Newer to US market, growing share in academic centers.

Percutaneous vs Paddle

Percutaneous leads (cylindrical, 8- or 16-contact) are placed through a Tuohy needle under fluoroscopic guidance, with or without paresthesia mapping. Paddle leads require laminotomy, are larger with broader contact arrays, and sit directly against the dura with less CSF attenuation between electrode and cord.

• Percutaneous: lower procedural morbidity, preferred for trial, reasonable durability in virgin epidural space, preferred when patient anatomy is straightforward. Migration rates meaningfully higher than paddle (5–15% reported across series).
• Paddle: lower migration (2–5%), more consistent stimulation at lower amplitudes (lower energy cost, longer battery), preferred for revisions, scarring from prior epidural instrumentation, or when percutaneous has failed. Higher procedural morbidity (laminotomy), slightly higher infection rate at the laminotomy incision.

The table below synthesizes consensus lead positioning for tonic (paresthesia-based) stimulation from Barolat's seminal mapping work, the NACC cervical and surgical-technique consensus papers, and current textbook recommendations. For 10 kHz high-frequency stimulation, anatomic placement at T9–T10 (occasionally T8–T11) without paresthesia mapping is the standard regardless of back-vs-leg distribution — one of the workflow advantages of the Nevro platform.

Remember that these are starting points for tonic-waveform paresthesia mapping, not prescriptions. Real-time paresthesia coverage confirms the correct level intraoperatively. For sub-perception and closed-loop waveforms the target tends to migrate slightly caudal (T10–T12 area) because of different fiber recruitment profiles.

Synthesized from Barolat mapping studies, NACC Cervical (2022) and Surgical Technique (2022) consensus papers, and SENZA trial protocols.

Positioning

Prone, pillow under the abdomen to flatten lumbar lordosis for thoracic entry (makes Tuohy trajectory less steep and dural puncture less likely). For cervical placement, head in pin fixation or foam positioner with slight flexion. Arms out at <90°, no brachial plexus stretch. A Jackson or Wilson frame is preferred to a flat table for obese patients.

Asleep vs Awake

The field has shifted. Ten years ago, paresthesia mapping during awake placement was standard. With the rise of 10 kHz and other paresthesia-free waveforms that use anatomic placement, asleep placement under GA has become routine — both more comfortable for the patient and more efficient for the OR. For awake tonic cases (classic paresthesia mapping), light sedation with cooperative patient is the sweet spot. Asleep placement requires intraoperative neurophysiological monitoring (SSEP or compound motor action potential [CMAP] mapping; the NACC surgical consensus provides the technical standard). This replaces awake paresthesia mapping as the confirmation of correct side-to-side and rostro-caudal position.

Percutaneous Technique Pearls

• Entry point typically two vertebral levels below the desired tip to allow a shallow epidural trajectory. For T8–T9 tip, enter at T10–T11. Steep trajectories increase dural puncture risk and make lead steering impossible.
• Paramedian entry, not midline. Enter the skin 1–2 cm lateral to the midline, aim the Tuohy toward the interlaminar space on the contralateral side. A paramedian trajectory gives you a shallow angle of entry to the epidural space.
• Loss of resistance with saline, not air. Air reduces the effectiveness of subsequent epidural stimulation if accidental subdural or intradural placement needs to be diagnosed with contrast, and introduces cord compression risk if inadvertent subdural air is injected.
• Advance under live fluoroscopy to confirm epidural space tracking. Lead should pass smoothly. Resistance implies extraspinal, subdural, or ligamentous tracking.