Heteroresistance among carbapenemase-producing (OXA-48) K. pneumoniae?

.

By Elena López-Camacho and Jesús Mingorance

Sometime ago we presented the meropenem E-test of a carbapenemase-producing K. pneumoniae (OXA-48) retrieved from a clinical sample in our institution Micro lab (What is going on with this Klebsiella?). Despite a well-defined halo around the strip several colonies grew within the halo. This finding triggered several questions:

• What do these inside-the-halo colonies mean?
• Are these colonies resistant to meropenem?
• Mutant, heteroresistant strains or simply an artifact?
• How  should MIC be read?

Elena López Camacho and Jesús Mingorance researchers at IDIPAZ have helped us to understand what is going on with this phenomenon:

.

.

A) Defining heteroresistance phenomenon and heteroresistant population

• Heteroresistance was suspected due to the presence of colonies in the E-test inhibition zone. These colonies might be the representation of a resistant subpopulation that can be formed by stable mutants with a genetic change (for example: point mutation on the antibiotic target gene) or a isogenic (non-mutant) sub-population.

.

• There is a simple way to distinguish between these two possibilities. It consists on the re-growth of colonies from the E-test inhibition zone, and replication of the E-test with this inoculum:

1. If the population inside the inhibition zone was formed by mutants, the new inhibition zone should be smaller.
2. On the contrary, if the same phenotype (presence of colonies in the inhibition zone) appears again, then the resistant subpopulation cannot be formed by mutants, but is a phenotypic phenomenon called heteroresistance.

.

• Heteroresistance is defined as the resistance to certain antibiotics expressed by an ISOGENIC SUB-POPULATION of a microbial population, that is generally considered to be susceptible to these antibiotics according to traditional in-vitro susceptibility testing. The heteroresistant population is a mixed population of drug-resistant and drug-sensitive cells in a single clinical specimen or isolate where the proportion of resistant organisms may not be explicable by the natural background mutation rate alone. For example, for a drug to which resistance arises by single point mutations on a target gene, a susceptible population will produce spontaneous resistant mutants with a frequency of 10^-8-10^-9 (one cell per 10⁸-10⁹ cells). On the case of a heteroresistant population, the proportion of resistant cells is higher (frequency about 10^-5–10^-6). Unless the organism exhibits an hypermutable phenotype this frequency can not be explicable by the natural background mutation rate alone.

.

• Therefore, heteroresistance is defined as the presence of a high number of resistant cells that are not mutants. Culturing the resistant cells reproduces the original mixed population.

.

B) Studying heteroresistance phenomenon: PAP curves

• Detection of heteroresistance is difficult and labour-intensive. Population analysis profiles (PAP-CURVES) are experiments designed to quantify the resistant sub-population for a series of antibiotic concentrations. Ten-fold serial dilutions of a stationary-phase culture are done and spread onto a complete series of antibiotic-containing plates and, after an incubation at the appropriate temperature for 24h, the number of colonies grown on each plate are counted. Results are represented on a semi-logarithmic graph with colony counts (resistant cfu/ml) or frequency (resistant cfu/ml relative to viable cfu/ml) on the vertical axis and drug concentration on the horizontal axis.

.

Model PAP-curve:

• In a resistant population the counts do not change when the antibiotic concentration increases
• In a wild-type, fully susceptible population the count abruptly decreases to 0 at extremely low antibiotic concentrations.
• However, in a heteroresistant population the count rapidly decreases to about 10⁵ cfu/ml at low antibiotic concentrations but then plateaus and maintains this value even at high antibiotic concentrations. The yellow band represents the area where the resistant cell count (or frequency) comes into the natural background mutation rate.

.

C) Analysis of carbapenem heteroresistance in K.pneumoniae OXA-48 (KPOXA-48): Carbapenem PAP curves of the two mayor KP-OXA-48 clones (ST405 and ST11)

• Carbapenem heteroresistance was suspected due to the presence of colonies in the E-test inhibition zone on most of the KP-OXA-48 isolates.

.

• We analyzed three ST405 isolates (two of them were from our OXA-48 outbreak and the other is from a Belgian hospital), three ST11 isolates (all from our OXA-48 outbreak) and two sporadic clones, one is an ST45 isolate and the other is a porin mutant that belongs to ST846. The porin mutant does not present inhibition zone in E-tests, while the ST11 (5588) isolate showed a clean zone. All the other isolates show colonies inside the E-test inhibition zone.

.

• The slopes of the curves vary among the different isolates, but at high antibiotic concentrations all the points were near or inside the yellow band and the analysis of these colonies revealed that they were porin mutants. Differences on the resistant subpopulations frequencies at low antibiotic concentrations can be due to differences in the level of expression of the different β-lactamase genes carried by these strains.

.

• PAP analysis showed that KP-OXA-48 do not present heteroresistance when grown in liquid medium. But colonies grown inside the zone of inhibition of the carbapenem E-test strips were not mutants.

.

• This phenotypic resistance phenomenon  seems to be bound to growth on a solid surface and might result from temporal variations in the expression levels of β-lactamase genes coupled to inoculum effects. The clinical relevance of this phenomenon and thus, if it should affect E-test reporting remains unknown. Further studies are needed.