The Clinical And Bioinformatic CTX-M Problem

We’ll get to what CTX-M is in a bit, but first some biology background. Let’s talk about some game theory carbapenem antibiotics (also referred to as carbapenems):

Before we get started, let’s do a quick biology review. A serious health problem are carbapenem-resistant enterobacteriaceae (‘CRE’)–E. coli and relatives that are resistant to all antibiotics that start with ceph-, cef-, or end with -cillin or -penem (the beta-lactam antibiotics). To make things worse, these bacteria are usually resistant to most or even all other classes of antibiotics.

In the U.S., the most common form of carbapenem resistance is encoded by the KPC gene (‘blaKPC’), which confers resistance to all beta-lactams. The one weakness blaKPC has is when it’s faced with an inhibitor, such as ampicillin-sulbactam: the sulbactam inhibits KPC, so the beta-lactam ampicillin can kill the host. So, you might ask, why not use inhibitor combinations like ampicillin-sulbactam or amoxicllin-clavulanic acid (augmentin)? The problem is that most strains with blaKPC also carry genes that protect the bacterium from these combinations (such as blaTEM).

So the last-line treatment is a newer beta-lactam with an inihibitor, such as ceftazidime-avibactam.

As I discussed in the cited post, KPC alleles (KPC-8, KPC-30, and KPC-32 for those of you keeping score at home) that confer resistance to ceftazidime-avibactam have evolved. That is a serious problem, but when most hospitals realize they have an isolate that carries the KPC carbapenemase, whether or not it confers resistance to ceftazidime-avibactam, institute severe infection control measures. This is a relatively rare event, and these measures can be taken.

But a recent paper describes another mechanism for ceftazidime-avibactam resistance: the alteration of the CTX-M-14 beta-lactamase, which was observed in a patient. CTX-M-14 usually confers resistance to ceftazidime, but two mutations that change two amino acids can result in resistance to ceftazidime-avibactam (one of the mutations by itself confers decreased susceptibility to ceftazidime-avibactam).

Unlike KPC, which is very rare (even large hospitals usually get a handful per year), CTX-M-14 is much more common. If hospitals were to implement strict infection control every time a CTX-M-14 (or more broadly, CTX-M*) organism were found, they might have to start putting people in the parking lot. So this isn’t good.

But there’s another problem I mentioned in the post title: the bioinformatic one. Many hospitals are using gene and genome sequencing to assist infection control and diagnosis. In other words, we can determine, within reason, what antibiotics to use from the DNA sequence.

However, the actual sequence for this novel ceftazidime-avibactam CTX-M resistance allele has never been deposited in a sequence depository (e.g., GenBank). That means that any number of resistance anti-microbial resistance gene databases (Resfinder, CARD, NCBI) will never correctly identify this sequence–instead, it will appear to be ‘just another CTX-M’ beta-lactamase, that doesn’t confer resistance to ceftazidime-avibactam.

This seems fine.


Seriously, it’s 2017, and journals, when they publish a paper on a clinical important sequence, whether a gene or a genome, need to make the authors submit the sequence to a public database. If it’s not in a database, for all intents, it does not exist. It can’t inform clinical care–which is the purpose of publishing the damn article. The editors really screwed this up.

One would think editors, who want people to cite their papers, would realize that a sequence deposition is a great way to get someone to read a paper about the sequence. Regardless, this sequence needs to get into a public repository, and soon.

This entry was posted in Antibiotics, Genomics, KPC. Bookmark the permalink.