Ann Robinson’s research reviews—5 November 2021

Interim results for sotrovimab in covid: too soon to call

Covid-19 has killed nearly five million people worldwide and is still raging. The quest for effective treatments continues, especially for older populations and those with underlying health conditions, who are at greatest risk of hospitalisation and death. In this phase III trial, high risk patients (those with at least one risk factor for disease progression such as being older than 55 years old, diabetes, obesity) with mild to moderate covid-19 who received a single infusion of the monoclonal antibody sotrovimab were less likely to be hospitalised for any reason or die within a month compared with those who received placebo (3/291 (1%) v 21/292 (7%)). There was no significant difference in adverse events (17% v 19%). The numbers are too small to make any great claims; only three people in total were hospitalised in the sotrovimab group, and uncommon adverse events may not have shown up. In theory, sotrovimab should be safe because it’s made from an antibody from a patient who recovered from covid-19 and targets an antigen on the virus (not the human host). This was only an interim analysis, and the trial is ongoing.

N Engl J Med doi:10.1056/NEJMoa2107934

Covid immunity waning in Israel

This study from Israel found evidence of waning immunity among vaccinated adults who had received two doses of the Pfizer covid-19 vaccine. The national database showed that people who had had the second dose when their age group was first eligible were more likely to have covid in the study period (11-21 July 2021) than those who had had the second dose two months later (rate ratio of around 1.6 for all age groups over 16). The rate ratio for severe disease was 1.8 in the over 60s, 2.2 in 40-59 year olds, and too small to calculate in the 16-39 year olds. Cases of covid-19 in Israel had all but disappeared in May, only to resurge by July as the dominant alpha variant was replaced by the delta variant (98% of cases). It’s likely that vaccine efficacy wanes over time and is less effective against the delta variant, but this study couldn’t distinguish between the two factors. Either way, these results led to a decision to offer a booster to anyone who had a second jab more than five months previously.

N Engl J Med doi:10.1056/NEJMoa2114228

A third covid jab provides an effective boost

Over 700 000 Israeli citizens with a median age of 52 years who were eligible for a booster dose of the Pfizer covid-19 vaccine more than five months after their second dose were matched with a control group. One week after the booster, vaccine effectiveness was estimated to be 93% for admission to hospital (231 events for those who had received just two doses versus 29 events among those who had received three doses), 92% for severe disease (157 v 17 events), and 81% for covid-19 related deaths (44 v 7 events). This study was observational, so the possibility of unmeasured confounding factors exists, and the possible harms of a third dose (such as myocarditis) weren’t explored.

An editorial discusses the dangers of global vaccine shortages, which could leave under-vaccinated countries susceptible to covid-19 and create the conditions for new variants that may come back to haunt the thrice-vaccinated countries.

Lancet doi:10.1016/S0140-6736(21)02249-2, doi:10.1016/S0140-6736(21)02388-6

Measure what needs to be measured: the pitfalls of surrogates

This important meta-analysis of 144 randomized clinical trials and over a million participants found that, just because treatments reduced the incidence of non-fatal myocardial infarction, they can’t be assumed to have reduced all-cause or cardiovascular mortality. Specifically, non-fatal myocardial infarction was not a surrogate for all-cause mortality in primary, secondary, mixed primary and secondary prevention, or revascularisation trials. The study highlights the danger of inferring too much from surrogate endpoints. It’s possible that, over time, improved diagnosis and prevention treatments meant that many non-fatal myocardial infarctions were no longer associated with increased mortality, but even studies from before 2000 (when high sensitivity troponin assays were introduced) failed to show a correlation between the effects of treatment on non-fatal myocardial infarction and reductions in all-cause and cardiovascular mortality.

JAMA Intern Med doi:10.1001/jamainternmed.2021.5726

Platelet-rich plasma injections for ankle osteoarthritis don’t work

Ultrasound guided platelet-rich plasma (PRP) injections—using a patient’s own platelets—into the ankle joint didn’t improve ankle symptoms and function in patients with ankle osteoarthritis over 26 weeks compared with placebo, according to this small Dutch trial. Ankle osteoarthritis may only affect 3.4% of the adult population, but it’s more common than hip and knee osteoarthritis among younger people, and there’s no effective, non-surgical treatment. Adverse effects were more common in the PRP group than the placebo group, who had an ultrasound guided intra-articular saline injection (13 v 8 events). Other PRP products may work better, although it’s hard to see why they would, and there was no information about other therapies that patients may have had (such as physiotherapy). On the basis of this study, I certainly wouldn’t let anyone inject PRP into my ankle.

JAMA doi:10.1001/jama.2021.16602

Ann Robinson is an NHS GP and health writer and broadcaster