Drug Shows Promise For Newborn Blindness

Does the story adequately quantify the benefits of the treatment/test/product/procedure?

Not Satisfactory

Consider this rating a partial credit. We were happy to see some results in absolute terms and the important subgroup result for babies with type 1 disease. But the article gets enough dings to withhold a satisfactory grade.

First are the qualitative summaries. “An injection of Avastin stopped blood-vessel growth.” That’s a bit vague for such a zinger. For one thing, one of Avastin’s apparent advantages was that it doesn’t stop vessel growth in the peripheral retina. For another, does this sentence mean Avastin cured all babies? (Retinopathy came back in 4 of 70.) Or does it mean Avastin universally stopped vessel growth at the outset? (The main endpoint of the study was recurrence of disease. Maybe the laser also stops vessel growth at first.)

And we think the story could set a neater table before comparing the benefits of Avastin and the laser therapy. It names 3 downsides of laser therapy, but efficacy (50% recurrence in zone I disease) wasn’t one. Since the study was designed to compare recurrence rates, we assume this downside of the laser is prominent. Another downside of the laser was given: some babies need several treatments. But, in comparison, we’re told that “in many cases” Avastin only needed one treatment. “Some do” means the same thing as “many don’t,” so a cleaner comparison would’ve been nicer.

Here’s a final aside, not of the same magnitude as the prior comments. In an ideal world, the article could’ve addressed the different types of retinopathy of prematurity. Indeed Avastin was superior in the overall cohort, especially in Zone I disease, and the article accurately quotes the editorial as suggesting Avastin should become the treatment of choice in Zone I. But absent is the fact that this study showed no significant benefit in Zone II disease, the larger group of enrolled babies. There was a trend towards benefit, but it wasn’t statistically significant. The message from the editorial is that Avastin is possibly superior in Zone II, but we’ll need more studies. This thread would’ve added some nice balance to the story’s conclusion that Avastin “far surpassed” laser therapy in treating this disease, and would allow them to be more specific about the 15 hospitals that stopped using the laser (for whom? all zones?).

Does the story adequately explain/quantify the harms of the intervention?

Not Satisfactory

Room for improvement. We give the article credit for noting that Avastin has limited safety information in this setting and that the investigator has noted no significant side effects in this cohort. It also gives some theoretical reasons to be optimistic. But what’s missing from the story is the key line from the study: it “was too small to assess safety.” You know the saying about absence of evidence and evidence of absence?

As noted, as readers we are unclear why Avastin is sparring with Lucentis. The article elevates Avastin because it “can’t easily travel outside the eye,” while Lucentis has a theoretical concern of systemic effects. That paragraph gave us several itches:

• The editorial tallies Avastin’s potential for systemic side effects as one of its disadvantages.
• Without knowing the role of Lucentis in therapy, it doesn’t matter much how it stacks up to Avastin hypothetically. Without more context, it seems like a straw comparison that serves to make Avastin look safe.
• The study states that a much larger trial would be needed to investigate Avastin’s potential for systemic side effects. That means the jury is still out, and systemic side effects are a live concern.

The article states that the 7 deaths in the study were unrelated to either treatment. We don’t see that conclusion in the publication or editorial. Perhaps it was based on an interview with the lead researcher? If so we’re unable to verify, since we can’t interpret the raw mortality information on the 7 cases in the study report ourselves. With that caveat in mind, we note that, to us, the study report seems to suggest that their study was too small to determine whether Avastin is associated with more deaths than is laser therapy. Since there were more deaths in the Avastin group (5 vs 2), we would’ve liked more clarity around this topic.