Doc — after digging into the numbers, I’m less worried about Simparica than the raw adverse-event reports make it look, but I would still treat the neurologic signal as real.
The cleanest way to understand it is this:
| Data source | What it shows | What it doesn’t show |
|---|---|---|
| FDA FOIA adverse-event reports through Sept. 2017 | 1,361 Simparica reports; 44 included death, 279 seizure, 560 tremor/shaking/ataxia | Risk per treated dog |
| U.S. controlled field trial | 315 Simparica-treated dogs; no seizure signal at labeled dosing; one transient ataxic episode | Very rare post-market reactions |
| EU/UK post-marketing pharmacovigilance | Neuro signs classified very rare: <1 treated animal per 10,000 | Exact individual risk for a particular dog |
| FDA current labeling | Tremor, ataxia and seizure are recognized post-marketing events | A numerical incidence rate |
The frightening FDA numbers
The frequently circulated dataset contained 1,361 reports involving Simparica. Within those reports:
- 44 deaths — 3.2%
- 279 seizures — 20.5%
- 560 shaking/tremor/ataxia events — 41.1%
- 91 other neurologic/cognitive events — 6.7%
Those figures are real report counts from the old FDA FOIA material. (PubMed Central (PMC))
But 44/1,361 does not mean 3.2% of dogs given Simparica died.
And 279/1,361 absolutely does not mean 20.5% of dogs taking Simparica seize.
The denominator is:
dogs for which somebody filed an adverse-event report
—not dogs taking Simparica.
It’s essentially asking, “Among people who called the complaint department, what percentage complained about seizures?” That’s an entirely different question from, “Among everybody who bought the product, how many seized?”
FDA explicitly warns against calculating risk or comparing products from its animal AER database because there is no exposure denominator, reports don’t prove causation, reporting rates vary, and underlying disease/concurrent drugs may be involved. (U.S. Food and Drug Administration)
What happens when we actually get a denominator?
This is the part I find much more useful.
European/UK regulators classify the post-marketing occurrence of tremor, ataxia and convulsions from Simparica as “very rare,” defined as fewer than 1 animal per 10,000 animals treated, including isolated reports. (Noah Compendium)
That puts the scale roughly here:
<0.01% reported frequency
rather than anything remotely resembling 20%.
And importantly, the product information says that in most reported neurologic cases, the signs are transient. (Noah Compendium)
That doesn’t mean “zero risk.” It means the best regulator-derived frequency category with an exposure denominator puts the event in the very rare range.
But I don’t dismiss the seizure signal
There’s good biological evidence that this isn’t merely noise in the reporting system.
In the Simparica margin-of-safety study, dogs given the normal maximum labeled dose (1×) did not show neurologic abnormalities. At 3×, tremors and ataxia occurred in several dogs. At 5×, seizures and tremors occurred. The dogs recovered without treatment. (Zoetis)
There’s an even more interesting case in the development program: one dog receiving about 3× the maximum labeled dose developed lethargy, anorexia, ataxia, tremors, disorientation, hypersalivation, proprioceptive abnormalities and absent menace after its third dose. Plasma sarolaner unexpectedly shot up, eventually producing a Cmax more than seven times the mean Cmax at the maximum labeled dose. The dog was ultimately euthanized. Investigators couldn’t determine why its sarolaner concentration suddenly became so high. (DailyMed)
That is much more persuasive to me than Facebook anecdotes.
It demonstrates a dose/exposure-related neurologic toxicity signal.
And pharmacologically that makes sense: isoxazolines interfere with arthropod ligand-gated chloride channels, with much greater selectivity for parasite receptors, but apparently not infinitely greater selectivity.
What about death?
This requires the most caution.
The old dataset’s 44 death-associated reports mean that death was an outcome mentioned in 44 reports where Simparica exposure was also reported. They do not establish that Simparica killed 44 dogs. (PubMed Central (PMC))
Dogs taking flea/tick products also die from hemangiosarcoma, GDV, CHF, trauma, epilepsy, toxins, renal disease, etc. Spontaneous pharmacovigilance deliberately collects these events because that’s how uncommon drug signals are detected.
Controlled field studies illustrate the problem beautifully. In one sarolaner study, severe events included deaths from pre-existing cardiac disease and euthanasia from severe arthritis, none judged treatment-related. (PubMed Central (PMC))
So I would not translate the post-marketing death reports into a drug-attributable mortality rate. There isn’t enough information to do that responsibly.
There’s another reassuring piece
In the pivotal U.S. Simparica field trial, 315 dogs received Simparica monthly for three treatments.
Reported adverse events were:
- vomiting: 0.95%
- diarrhea: 0.63%
- lethargy: 0.32%
- inappetence: 0%
One 8.6-year-old dog developed transient lethargy/ataxia/inappetence the day after Simparica plus ivermectin/pyrantel; the signs resolved the following day. (DailyMed)
A trial of 315 dogs obviously cannot tell you much about an event occurring around 1 in 10,000. You’d expect essentially zero such cases in a study that size.
And that’s exactly why the neurologic warning emerged post-marketing rather than screaming out of the approval trial.
So would I give it to my dog after reading all this?
Yes — with one important qualifier.
For a healthy dog with no seizure disorder, unexplained tremors, episodic ataxia, vestibular weirdness or previous isoxazoline reaction, I’d be comfortable giving Simparica.
The combination of the evidence tells me:
The adverse effect is real.
The neurologic mechanism is plausible.
Susceptible individuals apparently exist.
But the population-level risk appears quite small.
The EU regulatory classification of <1/10,000 treated animals carries enormously more weight with me than “20.5% of Simparica adverse-event reports involved seizures.” (Noah Compendium)
Where I’d change course is a dog with a seizure history. FDA specifically says seizures have occurred even without previous seizure history, but recommends caution in animals with existing seizures or neurologic disorders. (DailyMed)
For that dog, I’d probably say:
Why deliberately expose a neurologically vulnerable animal to an agent with a known neurologic class effect when I have other parasite-control strategies?
And one subtle change from what I told you earlier: I wouldn’t say the evidence shows Simparica is completely interchangeable with NexGard, Bravecto and Credelio from a safety standpoint. They share a class warning, yes, but spontaneous-report counts aren’t suitable for proving that their true risks are identical. FDA specifically warns against making those cross-product comparisons from AER counts. (U.S. Food and Drug Administration)
So my bottom line, Jay-bird:
I would use Simparica in my own neurologically normal dog. I would avoid the entire isoxazoline class in my own epileptic or neurologically questionable dog if a practical non-isoxazoline strategy could give me adequate parasite protection.
And I’d tell an owner the risk roughly like this: “There is a genuine seizure/tremor warning. It isn’t imaginary. But the scary online percentages are percentages of adverse-event reports, not percentages of treated dogs; post-marketing regulators classify these neurologic events as fewer than 1 in 10,000 treated animals.”
That is a very different informed-consent conversation. (Noah Compendium)
