Complex Regional Pain Syndrome: Bisphosphonate, Ketamine, or Gabapentinoid First?
A positive three-phase bone scan four months into complex regional pain syndrome points toward a specific mechanism — and toward the one drug class whose own trials selected for exactly that finding.
Walter S., a 58-year-old man who builds simple wooden birdhouses for each of his six grandchildren every winter, fractured his distal radius falling off a ladder while cleaning gutters, treated with closed reduction and casting rather than surgery. Four months later the cast is long off, but his hand and forearm remain swollen, shiny, and noticeably warmer to the touch than the other side, with allodynia severe enough that a bedsheet brushing his wrist at night wakes him. He meets the Budapest clinical criteria for CRPS Type I, and a three-phase bone scan ordered to characterize the process shows markedly increased periarticular uptake on the delayed images — a finding present in a meaningful minority of early CRPS cases and specifically associated, in the literature that guided the drugs now being discussed, with a bone-turnover-driven phase of the disease rather than its later, more fibrotic stage.
That imaging finding turns out to matter more than it might look like it should, because the strongest trial evidence for each of the three options on the table was generated in populations selected somewhat differently. Bisphosphonates carry the most encouraging randomized evidence of any pharmacologic option in CRPS: Varenna and colleagues' 2013 trial of intravenous neridronate enrolled 82 patients within the first year of onset and found substantial pain reduction in a population whose disease — like Walter's — was still in an active phase rather than long-fixed. That result has not translated cleanly. The multinational program that followed tested lower cumulative doses and failed to separate from placebo, and neridronate remains investigational in the United States, licensed for CRPS only in Italy; the FDA has granted it orphan, fast-track, and breakthrough-therapy designations, but the only route to the drug for an American patient today runs through a trial. The phase 3 study now enrolling asks for warm-phase CRPS-1 of six months' duration or less with a positive three-phase bone scan — which is to say it asks for Walter, four months in with exactly that scan. Ketamine infusion's evidence, from Schwartzman and colleagues' and Sigtermans and colleagues' separate double-blind trials, showed real pain reduction in CRPS as well, but both trials enrolled broadly by clinical diagnosis and symptom duration without requiring or even reporting bone-scan positivity, so their populations don't specifically confirm benefit is concentrated in patients like him. Gabapentin's own dedicated CRPS trial, from van de Vusse and colleagues, found a real reduction in sensory symptoms but no significant effect on pain intensity or motor findings — the weakest of the three results on the outcome that actually brought Walter into clinic.
Pain clinic consultation, four months post-fracture
His bone scan is doing real work here, not just confirming a diagnosis he already had clinically. Varenna and colleagues' 2013 trial of intravenous neridronate enrolled patients within roughly the first year of CRPS onset and found substantial pain reduction sustained past the infusion course itself — a population defined by duration that overlaps closely with an active, imaging-positive bone-turnover phase like his. This is the one option where the trial population and this patient aren't merely similar in diagnosis but in the specific process his imaging shows. I have to be straight about what that gets him, though: neridronate isn't approved here. It's licensed for CRPS in Italy and investigational everywhere else, and the multinational trials at lower doses didn't beat placebo. What I'm proposing is enrollment in the phase 3 study, whose entry criteria read like a description of him — warm phase, under six months, positive three-phase scan.
The population match is a fair point, and I'm not disputing that his scan shows an active process — it does.
But what's actually disabling him is the allodynia, a sheet touching his wrist waking him at night, and that's a central sensitization phenomenon that ketamine's NMDA blockade addresses directly. Schwartzman and colleagues' and Sigtermans and colleagues' double-blind ketamine trials both found real pain reduction in CRPS without requiring bone-scan positivity as an entry criterion at all — their evidence doesn't need his imaging finding to apply to him, it applies on the strength of his allodynia and his diagnosis alone.
Both of you are arguing about which infusion protocol to arrange, and either one is going to take time to schedule and requires monitoring neither of us can provide in this clinic today. Van de Vusse and colleagues' gabapentin trial in CRPS found a real reduction in sensory symptoms, even though it didn't move pain intensity or motor findings significantly — weaker than either option on the table, plainly.
I'd start gabapentin tonight while the trial referral gets arranged, not instead of it. He doesn't have to choose between the strongest evidence and something starting today if we're honest that this isn't really a competition — it's a bridge.
Gabapentin started same-day and titrated over two weeks while trial screening for neridronate is arranged, both proceeding in parallel rather than sequentially. If he declines enrollment or screens out, IV pamidronate stands as the bisphosphonate actually available to him here. Sleep and allodynia severity tracked weekly as the primary functional outcome.
If allodynia remains severe once a bisphosphonate course is completed, ketamine infusion moves from held-in-reserve to active consideration without needing to revisit today's reasoning.