Dr. Mike Selby’s talk showcases how prone lateral surgery becomes a powerful revision tool for TLIF/PLIF non‑union and infection, letting the surgeon address posterior hardware and the anterior column in a single position while minimizing destabilization and morbidity.
Intro and concept of prone lateral
- His recommendation: gain solid comfort with at least ~20 traditional lateral cases (or equivalent high‑volume mentored experience) before attempting prone lateral, because the anatomical understanding transfers directly but the positioning and workflow are more complex.
-
Prone lateral particularly shines in revision settings: it avoids an extra flip, allows simultaneous or staged posterior and lateral work, and provides a favorable corridor as the peritoneal contents fall anteriorly in the prone position.
General remarks on revision and cage removal
-
Revising posterior non‑union around interbody cages is described as “hating to see CSF,” because posterior extraction across a scarred dura carries high risk of durotomy, infection, and junctional problems.
-
By contrast, anterior/lateral revision lets the surgeon place large‑footprint cages, bypass compromised endplates, and remove posteriorly placed implants through a safer corridor, as long as vascular/visceral anatomy is respected.
-
He shows extracted TLIF cages (PEEK), noting that PEEK breaks readily under osteotome—an advantage when fragmenting cages—while warning not to ruin standard osteotomes or hooks on titanium devices; dedicated extraction tools should be used.
Case 1: 80‑year‑old with L3–S1 fusion, L3–4 non‑union and proximal ASD
-
An 80‑year‑old woman with four prior spine surgeries (by others), fused from L3–S1, presents with severe back pain, inability to stand straight, and adjacent segment disease at L1–2 and L2–3 plus suspected non‑union at L3–4.
-
Flexion–extension films are equivocal; a SPECT bone scan “lights up” at L3–4, confirming this as the painful motion segment, while L4–S1 appears solid.
-
MRI shows proximal central/foraminal stenosis but no major compression at L3–4; the plan is to address the non‑union and extend proximally using prone lateral to avoid a big anterior approach.
Technique and outcome
-
Posteriorly, he and a neurosurgical colleague remove old hardware and prepare for extension; then, without flipping, they rotate to the prone lateral corridor using a three‑arm retractor with a glove over it to hold back retroperitoneal contents.
-
Under lateral visualization, they free and remove the mobile TLIF cage at L3–4 using osteotomes and cage hooks, demonstrating that what looks “fused” on X‑ray can be a fibrous non‑union when directly inspected.
-
He then performs additional prone laterals at L1–2 and L2–3 and extends fixation up, leaving the patient fused in slight flatback but acceptable for an 80‑year‑old; follow‑up imaging shows a continuous construct with resolved motion at L3–4, and she is functioning well.
Case 2: complex infected construct with PJK, cage infection and flatback
-
A man in his late 50s/early 60s with seronegative arthropathy initially underwent multilevel fusion and later extension to T11 for adjacent segment disease, then developed wound drainage, a deep infection with epidural abscess (Staph epidermidis), prolonged IV and oral antibiotics, screw loosening, PJK, and flatback.
-
When he presents to Selby, he has pain at the thoracolumbar junction, positive sagittal balance (PI 45°, LL 17°), coronal imbalance, and radiologic evidence of infection and non‑union, but is not septic.
-
A gallium scan localizes active infection to the most recently operated level (the PJK extension level), implicating those cages as the main focus rather than the entire construct.
Strategy: “orthopaedic” infected non‑union logic
-
Selby and discussant Andrew Toogood talk through options: implant holiday vs attempts at salvage with wound VACs and staged washouts, noting how hard it is to decide when posterior implants and cages must be removed versus retained.
-
Drawing on long‑bone principles, Selby emphasizes that infection clears better with stability: in tibial infected non‑unions one removes the nail, debrides, and adds external fixation; he applies the same logic with a new stable construct over a debrided front.
Execution with prone lateral, navigation, and bioglass
-
In a single anesthetic but conceptually staged fashion, they:
-
Posteriorly remove the infected T11–L1 segment of hardware, debride thoroughly, and take new cultures.
-
Re‑prep and re‑drape for prone lateral; via lateral corridor, remove the infected L1–2 (and/or adjacent) cage(s) and replace with expandable titanium cages packed with bioglass plus antibiotic powder, avoiding allograft in this infected bed.
-
Use an O‑arm for navigation to place new screws in the upper thoracic spine and pelvis in a “virgin” plane, then drop pre‑contoured rods from T6 to the pelvis, effectively bridging the debrided, previously infected region with a fresh construct.
-
Case 3: distal non‑union after adult Scheuermann kyphosis correction
-
A 43‑year‑old woman with thoracolumbar Scheuermann‑type deformity, thoracic pain, and chronic low back pain (no neurologic deficit) undergoes a primary T4–L3 posterior fusion with multilevel Ponte osteotomies, cement augmentation, sublaminar bands, and an outrigger rod; early EOS shows good correction without overcorrection.
-
She does well for about two years, then returns with new back pain; imaging reveals a distal screw break and incomplete facet fusion at the lowest instrumented level—essentially a focal distal non‑union.
Prone lateral “distal rescue” and prophylaxis concept
-
Rather than extending to another distal level, Selby uses prone lateral to insert an interbody at the non‑union level and, in the same sitting, reopens the distal posterior segment, lifts the rod ends, and upsizes/replaces the broken screw, adding support while preserving the current distal level.
-
This case convinces him that prone lateral is so effective for distal junction non‑unions that he now often places a distal prone lateral cage at the index surgery in patients he considers high risk, as a prophylactic measure to reinforce the base of the construct.
-
Long‑term EOS images show maintained correction and good alignment; clinically, she is very satisfied. During the same anesthetic, he also revises an asymptomatic but osteolytic M6 cervical disc replacement to fusion, highlighting how prone lateral can be combined with other procedures under one anesthesia.
Practical pearls and closing points
-
Selby underscores key tactics:
-
Always know prior implants and bone quality; obtain original drivers when possible.
-
Use navigation and pre‑contoured rods for long infected revisions.
-
Respect soft tissues, avoid burning future revision planes, and do not hesitate to stage when needed.
-
-
All agree there is no single right answer for when to remove vs retain infected cages and screws; timing, causative organism, systemic illness, mechanical stability, and localization scans (e.g., gallium) all shape the decision.
-
He concludes that prone lateral will likely become a standard tool in the next 5–10 years for revision of TLIF/PLIF non‑unions and infections, because it saves time, avoids flipping, and enables genuine simultaneous anterior column and posterior work in complex adult deformity and revision practice.