#apaperaday: Adverse events associated with eteplirsen: A disproportionality analysis using the 2016–2023 FAERS data

In today’s #apaperaday, Prof. Aartsma-Rus reads and comments on the paper titled: Adverse events associated with eteplirsen: A disproportionality analysis using the 2016–2023 FAERS data

Today’s pick is from Heliyon by Dai et al on an analysis of the side effect database FAERS (FDA adverse event reporting system) for reported side effects for eteplirsen use in Duchenne. DOI: 10.1016/j.heliyon.2024.e33417

Eteplirsen has been approved for Duchenne by the FDA in 2016. Eteplirsen is a phosphorodiamidate morpholino oligomer (PMO) antisense oligonucleotide (ASO) that induces exon 51 skipping in Duchenne and allows patients to make a partially functional dystrophin.

As it has been approved for almost 8 years now, real-world evidence collection allows assessing long-term side effects in a real-world setting. Notably these will not be apparent in clinical trials as these are shorter and only involve a selected group of patients.

Here authors looked into the FAERS database for side effects reported for eteplirsen-treated patients between Q3 2016 and Q4 2023. Side effects were reported for 1480 patients with 3107 side effects. Most patients were from the USA and most side effects were reported for males.

Authors discuss that this makes sense as the treatment was approved in the USA and most Duchenne patients are males. I would even say that virtually all patients eligible for exon 51 skipping are males—manifesting carriers with an exon 51 skippable variant I would not treat.

The reason is carriers will also have a normal dystrophin transcript. While eteplirsen will allow them to produce a partially functional dystrophin if exon 51 is skipped in mutated transcripts, if exon 51 is skipped from normal transcripts, this disrupts dystrophin production.

So you will get more partially functional dystrophin at the cost of less fully functional dystrophin. Seems like a bad trade-off to me. But back to the paper: 13% of the reported side effects resulted in hospitalization, 2.7% to death, and 0.6% to life-threatening situations.

The challenges with the side effects reported in the self-reporting system are that context is often missing. With the deaths, we do not know whether the treatment immediately caused death, or whether patients succumbed to their disease, or whether it was a combination.

If older patients are fragile and heart function is not optimal, infusion may result in acute problems and then things may spiral out of control with sad consequences. Back to the paper again: the vast majority of side effects reported were due to the infusion.

This involved administration site problems, infections, vascular problems, etc. Respiratory disorders were also reported, but again this may be due to the disease or a combination of treatment procedure and disease. Adverse events generally happened 903 days after the 1st treatment.

Authors discuss the importance of following up and monitoring adverse events for new treatments—especially those receiving accelerated approval. Catheter problems were the most reported side effect—note that this is due to the fact that intravenous infusion is needed.

It is not caused by eteplirsen per se, but by the procedure of administering it. Most side effects fortunately were mild. Authors also mention more severe side effects such as sarcoma, fat embolism, cardiomyopathy, and cardiac death.

The sarcoma may be a chance finding. Fat embolism is more common in Duchenne patients because they have a higher likelihood of long bone fractures. Cardiomyopathy is also part of Duchenne pathology. So again I do not think this was caused by eteplirsen.

The sudden cardiac death may have been the result of the infusion, but likely in an individual who already was fragile and had poor heart function. As context is missing we can only speculate. Authors point out that knowing what to expect means we can monitor and manage better.

I fully agree. Authors also stress the limitation of the database as context is generally missing, and often not all data are available about age, race, etc. (thus making it difficult to see whether some side effects are more common in specific age groups or races).

Furthermore, we do not know about side effects not reported into the database, as it is a voluntary action and not a mandatory one. Still, I like that the authors did the analysis—they used higher-level statistics. I do not have the expertise to see if they did that correctly.