RR:C19 Evidence Scale rating by reviewer:
***************************************
Review: Immunocompromised patients infected with SARS-CoV-2 represent several challenges. They can have more severe outcomes, and become persistently infected, with long-term shedding of viral particles. If the patients are treated with antivirals, over time there could be a selection for variants of the virus that are resistant. These variants can then spread through continual viral shedding of the patients. The purpose of the paper by Nooruzzaman et. al paper was to document the development of antiviral resistance in SARS-CoV-2 genomes, over time, in immunocompromised patients.
To do this, they examined a set of 15 immunocompromised patients that had persistent SARS-CoV-2 infections and had been treated with antivirals. SARS-CoV-2 viral genomes were extracted and sequenced from all patients, including multiple time points to study variation over time. Two commonly used antivirals are remdesivir [Veklury™], which targets the activity of the viral RNA-dependent RNA polymerase (nsp12), and nirmatrelvir-ritonavir [Paxlovid™] which targets the viral nsp5 protease. Nine of the fifteen patients had mutations in the nsp12 gene, and four had mutations in the nsp5 gene. There were some viruses isolated with mutations in both nsp5 and nsp12 genes, making them less sensitive to both antivirals, and these viruses were found to be as transmissible as wild-type SARS-CoV-2 viruses in a hamster model.
The authors were successful in their demonstration of how antiviral resistance can emerge and in showing that these resistance variants can be spread in the environment potentially infecting others and reducing the effectiveness of antiviral treatments. There is a clear need for a discussion of how best to treat immunocompromised patients with SARS-CoV-2 infections. If antivirals are not effective for these patients and can in fact lead to resistance, alternatives should be considered.
I do not have any major criticism of this work – the approach is clear and the methods are described well. As a minor point, I am curious as to whether directly sequencing the viral genomic RNA might be considered as an alternative to the more traditional RNAseq, where a DNA copy of the viral genome is made, then chopped into small pieces and sequenced and stitched back together. Using third generation single-molecule sequencing, such as Oxford Nanopore flow cells, it is possible to directly sequence the viral genomic RNA, with reads potentially covering the entire length of the viral genome, and this was reported for the SARS-CoV-2 genome (PMID: 32330414). As I see it, there are several advantages to this new technology: first, it is easier and potentially less expensive; second, it is possible to quantitate variants with single molecule reads, and the long reads will allow for quantitation of alternative splicing and different transcript lengths, for example. Further, RNA base modifications can be routinely detected in viruses [PMID: 28278189], and more than 50 years ago it was proposed that 5mA could play a role in viral infection [PMID: 166375]. There are more than 170 different possible base-modifications in RNA, and it’s likely that viruses use some of these (PMID: 36266445). When Oxford Nanopore flow cells were first made commercially available, more than a decade ago, their error rates were quite high, and yields low. Now some of the error rates are known to be due to base-modifications, and with slow and steady improvement to the technology, the error rates are becoming much lower and the read lengths quite long.
Creative Commons Attribution 4.0 International License (CC-BY 4.0)
