Serotonin Syndrome or Neuroleptic Malignant Syndrome: An Overlapping Exposure
A post-surgical patient develops fever, rigidity, and altered mental status after exposure to both a serotonergic drug and a dopamine-blocking one. His exam doesn't cleanly point to either syndrome, and the two syndromes require different treatments.
Daniel O., a 45-year-old long-haul truck driver who spends most weekends restoring a 1974 camper van in his driveway, is two days out from an elective L4-L5 laminectomy for a herniated disc that had kept him off the road for three months. He has taken fluoxetine 40mg daily for treatment-resistant depression for six years, stable enough that his psychiatrist hadn't adjusted the dose since before his first grandchild was born, and reports no other regular medication. His nausea on the evening of surgery — significant enough that he'd vomited twice — was treated with prochlorperazine. Post-op day one was otherwise unremarkable; his pain was managed with scheduled acetaminophen until that fell short by early afternoon, when he was given his first dose of tramadol. By the morning of post-op day two he is agitated, tremulous, and drenched in sweat, with a temperature of 39.4°C that wasn't present twelve hours earlier.
Both agents on his post-op medication list are individually capable of producing exactly the picture in front of the team, through mechanisms that point in different directions. Tramadol is not just an opioid — it inhibits serotonin and norepinephrine reuptake in its own right, on top of its active metabolite's mu-agonism, which is precisely why its labeling carries a specific warning against combining it with an SSRI; added to six years of steady fluoxetine, whose active metabolite lingers for weeks even after the parent drug is stopped, it is a textbook serotonin-syndrome precipitant. Prochlorperazine, by contrast, is a potent D2 antagonist with a genuine, if less commonly cited, case-report association with NMS. His exam sits in the genuinely contested space between the two: reflexes are brisk at the ankles with a few beats of inducible clonus, arguing for serotonin syndrome's hyperexcitable picture, but his forearms show a rigidity his nurse describes as "like bending a lead pipe" — the textbook NMS finding rather than the myoclonic jerking more typical of serotonin excess. His creatine kinase, drawn an hour ago, is already climbing, and the two syndromes' natural histories diverge sharply from here if nothing changes: serotonin syndrome tends to worsen over hours without treatment, NMS over days. He is exactly at the point on that timeline where the two curves haven't yet separated enough to read.
On the ward, twelve hours into an ambiguous fever
I'd call this serotonin syndrome and treat it as such right now. He meets the Hunter criteria — Dunkley's 2003 decision rules — on the inducible-clonus limb: inducible clonus plus agitation, with diaphoresis on top of it, in a patient on a serotonergic agent — and the exposure history couldn't be cleaner: tramadol inhibits serotonin reuptake in its own right, independent of its opioid activity, and its own labeling specifically warns against combining it with an SSRI. Six years of steady fluoxetine means his active metabolite was never really 'off' to begin with; tramadol on post-op day one was the second serotonergic hit, not an isolated exposure. Stop the tramadol, start lorazepam for the agitation and hyperthermia, and reserve cyproheptadine if he doesn't turn around over the next few hours. I wouldn't reach for dantrolene or bromocriptine here — bromocriptine has its own case reports of worsening serotonergic toxicity, which is the last thing I want to hand a patient whose actual diagnosis I think is serotonin syndrome.
The reflex finding doesn't settle it for me the way it's settling it for you. What I keep coming back to is the rigidity — "like bending a lead pipe" is the nurse's own description, and that's the textbook NMS finding, not the myoclonic jerking serotonin toxicity usually produces. His CK was already climbing an hour ago, and prochlorperazine is a potent D2 antagonist with a real, if smaller, literature behind it as an NMS trigger. If this is NMS and we spend the next several hours treating it as serotonin syndrome, he's accumulating muscle breakdown and autonomic instability the whole time benzodiazepines alone aren't going to touch.
And calling him Hunter-positive cuts the wrong way for you. He would also satisfy Dunkley's rules on the hypertonia limb — rigidity plus a temperature over 38 plus inducible clonus — a rule he can meet on findings that sit just as comfortably in NMS. Criteria derived in serotonergic overdose patients don't discriminate between two syndromes when only one of them was ever in the sample.
You're both arguing from real, individually defensible readings of the same exam, which is exactly why I don't think either of you should get the specific antidote yet. Stop both drugs — tramadol and prochlorperazine — right now; neither of you disputes that much. Start aggressive benzodiazepines for the agitation and hyperthermia; that helps regardless of which of you turns out to be right. Boyer and Shannon's New England Journal review makes benzodiazepines the first move for agitation and hyperthermia in serotonin toxicity, and they are equally standard supportive care in NMS — which is exactly the point. It's the one intervention that doesn't require winning the argument first. Recheck him in two hours. If the rigidity is worsening and the CK keeps climbing while he stays lead-pipe stiff, that's NMS declaring itself, and dantrolene plus bromocriptine make sense then. If the clonus spreads and he stays hyperreflexic with the rigidity static or improving, that's serotonin syndrome, and cyproheptadine is the right next step. Committing to either drug right now, before the picture separates, risks giving the wrong one to the wrong syndrome when waiting two hours costs us almost nothing that benzodiazepines and stopping both agents haven't already covered.
Agreed within the hour: both tramadol and prochlorperazine stopped, IV lorazepam started and titrated to agitation and vital signs, active cooling measures begun, and a two-hour bedside reassessment scheduled with the specific branch point named in advance — worsening rigidity and climbing CK toward dantrolene and bromocriptine, spreading clonus and persistent hyperreflexia toward cyproheptadine.
Not agreed, and left as an open instruction for whoever is at the two-hour recheck: how much weight a CK value alone should carry if it keeps rising on both possible readings, since CK elevation is well described in prolonged agitation and rigidity from either syndrome and isn't in itself the tie-breaker the neurologist would like it to be. His fluoxetine was held pending psychiatric input on whether restarting it later needs a slower reintroduction or a switch to a non-serotonergic agent altogether — that conversation deferred to the outpatient team, not resolved tonight.