Has the bell tolled for redback antivenom?

"Latrodectus hasseltii close" by Toby Hudson - Own work. Licensed under CC BY-SA 3.0 via Wikimedia Commons.


Redback spider bites are of decent medical importance, if you live in Australia or any of the places they’ve been carried to by humans. Their venom inflicts you with pain that can be local, regional, or generalized. It can also cause systemic effects such as nausea, vomiting, headache, or malaise. Due to this, Australian researchers have carried out multiple studies on the utility of redback antivenom for treating the effects of bites.

This one, recently e-published in Annals of Emergency Medicine, is a follow-up RCT to a prior study. This study compared antivenom to placebo.  That prior study compared IM to IV antivenom, and was also an RCT. It found statistically no difference between the two routes, so they used IV dosing for the second study.

Methodologically this paper is sound. They describe their power calculation, and the blinding is pretty impressive. They even used an independent analyst to check the data before the blinding was removed. Sadly, they had to close the trial 16 patients early, due to lack of funding. They simply couldn’t afford to supply the 20 study hospitals with another set of test vials.

One drawback is that the spider wasn’t identified nearly 1/4 of the time. This is less of a concern than in, say, my patient population, where the MRSA spider is endemic. Aussie clinicians are fairly familiar with redback envenomations, and the other venomous spiders have distinct envenomation syndromes.

For their primary outcome of improvement in pain at 2 hours, there was a 10.7% difference between placebo and antivenom (23 vs 34%). Their second primary outcome was resolution in systemic effects at 2 hours, and there was only a 4% difference (22% placebo, 26% antivenom). Neither of those outcomes reached statistical significance. Secondary outcomes were improvement in pain at 4 hours and 24 hours, rescue treatments (second doses of unblinded antivenom or opiates) as well as use of opioids after discharge, followup medical care, and serum sickness. There was no significant difference in pain measurements at 4 and 24 hours, use of rescue treatments, or followup medical care. There were only 4 acute hypersensitivity reactions, all in the antivenom group. Interestingly, there were 5 cases of serum sickness in the placebo group (even after those who got rescue antivenom were removed from the count).

Thus, for Redback spider antivenom the NNH is 25, while the NNT is only 10. For pain secondary to redback bites, antivenom doesn’t work, but pain meds don’t seem to either. The authors note that more RCTs need to be directed towards identifying treatments that do work. Isbister himself has said that this has changed his practice to the point that he won’t use redback antivenom anymore.

Since latrodectus species are present on almost all continents (sorry Antarctica), and there is a fair amount of cross-reactivity among multiple species, should we apply this to all latrodectus spiders? In a word, no. Redbacks (L hasselti) cause pain, but aren’t as toxic as North American (L mactans and L hespersus) or European (L tredecimguttatus) spiders. Reports of death from redbacks are incredibly rare, and treatment is directed towards simply treating the pain. So while not using antivenom to treat pain is prudent, current recommendations for the US black widow antivenom are to prevent untoward outcomes from severe envenomations in the young and those with comorbidities. It is hard to power a study to show benefit in such rare events. Antivenom isn’t cheap, and it isn’t without risks. It’s basically the tPA of the toxicology realm. And it probably has just as much controversy.

It looks like it is time for @precordialthump to update his CCC on Australian antivenoms.

Randomized Controlled Trial of Intravenous Antivenom Versus Placebo for Latrodectism: The Second Redback Antivenom Evaluation (RAVE-II) Study.


Leave a Reply

Your email address will not be published. Required fields are marked *