The microbroth growth inhibition assay described in the present study correlated well with the results of the agar overlay spot method. Lytic phages were detected as early as 6 h of incubation using high phage (5 × 109 PFU/ml) and bacteria (5 × 108 CFU/ml) concentrations using a growth inhibition cutoff of ≥ 31% with 100%/100% specificity/sensitivity for differentiating phages with or without lysis. Continuous monitoring of OD630 of a low bacterial inoculum (5 × 106 CFU/ml) in the presence of a high concentration of phages (5 × 109 PFU/ml) for 6 h and calculation of growth inhibition based on the AUGC could help to differentiate phages with semi/transparent lysis from phages with turbid lysis with 100%/86% specificity/sensitivity using a ≥ 61% growth inhibition cutoff. Alternatively, lytic vs. non-lytic phages and semi/transparent vs. turbid lysis could be differentiated after 24 h of incubation using a high phage (5 × 109 PFU/ml) and bacteria (5 × 108 CFU/ml) concentrations with similar sensitivity and specificity.

In comparison to the classic overlay agar method, the microbroth assay can offer additional information on the phage–host interaction, including phage–host range, pathogenicity, and development of potential phage resistance at different phage and bacterial concentrations, in a high-throughput manner. Using the overlay agar method, normally a single phage concentration is applied in a bacterial lawn with a certain inoculum and the capacity of phage–host pairs is limited due to the space of the petri dish. In contrast, using a microtiter host-range assay users can test up to 48 different phages at a single phage concentration with one host, or explore various concentrations of phage–bacteria pairs, all within the format of a 96-well plate. The format can easily be adapted to a larger microplate, e.g., a 384-well plate, where more conditions can be tested. If kinetics could not be performed, then the lytic activity of multiple phages and bacteria can be tested in many 96-well microtitration plates with measurement of OD only at 6 h or/and 24 h calculating growth inhibition. Apart from that, human errors with subjective lysis estimations can be avoided because results are interpreted and calculated automatically using the plate reader and software. Finally, among the advantages of the microbroth method is the detection of potential false-negative results in the overlay method due to the development of phage resistance at later time points during incubation, which can be detected when kinetics are performed [15]. According to this assumption, the traditional agar overlay method approach for evaluating the phage host range may be replaced with the proposed microbroth plate method [9].

Due to the dynamic nature of plaque formation, variations in the phage latent period, burst size, diffusion rate, and host growth can all have an impact on the size and visibility of the plaque [16]. Growth curves of bacteria in the presence of phages using a 96-well plate is included among methods with potential for scalable high-throughput quantification [17]. Growing bacteria in an automated spectrophotometer produces metrics that are proxies for bacterial cell densities (e.g., optical density) at predetermined time intervals. Generating growth curves of bacteria in the presence of phages allows for observation of the effect of phages on bacterial population density over time [18, 19]. As described previously [9], phage concentration-dependent effects were also observed in our study, where phages with lower titers resulted in a decreased growth inhibition in comparison with higher titers. Growth inhibition at high phage concentrations of 5 × 109 PFU/ml correlated well with the agar overlay method where the initial bacterial inoculum in spot assay is 5 × 108 CFU/ml and the phage stock that inoculated in the bacterial lawn is usually > 1010 PFU/ml. Moreover, in a previous study, growth inhibition was found in a liquid-based assay while no lysis was observed in the spot assay [9]. We assume that this phenomenon was observed due to the higher phage concentrations used in the liquid assay compared to the standard spot assay. Another assumption is that due to the emergence of resistance in traditional spot assay, hypothetical formed plaques were not visible at the time of observation, while in liquid-based assays the continuous measurement of OD could predict growth inhibition before the emergence of resistance.

Compared to the conventional spot assay, the microbroth assay offers many advantages in evaluating the phage host range, like optimization by testing several phage and bacteria concentrations in order to correlate with in vivo outcome. In a mouse model of P. aeruginosa infection, spot tests showed a tendency to overestimate phage virulence [20]. Similarly, phage growth rate in liquid culture was found to be a reliable indicator of phage in vivo efficacy [21]. Additionally, in an in vitro study comparing microtiter liquid-based culture with spot assay for phage host range and virulence, results have shown that the two methods were in agreement with 6% false-negative results between the two methods [9]. According to our findings, which are in agreement with previous studies, microtiter plate assay was able to provide reliable results with high sensitivity and specificity for K. pneumoniae host range as well as virulence at high phage concentrations tested in a high-throughput format with reliable results. This method can minimize human error by automatically calculating the growth curves and results tend to be more objective rather than subjective.

Comparison of growth inhibition did not show any significant differences between the two inocula when actual phage concentrations were used as opposed to when MOI was used, indicating that actual (PFU/ml) rather than relative (MOI) phage concentrations may be pharmacodynamically important. From pharmacological and clinical perspective, the PFU/ml is commonly used to describe phage pharmacokinetics whereas the actual MOI is not known as the bacterial concentrations at the site of infection are usually not known. MOI was suggested not to be used as the sole dosing descriptor during phage therapy (https://doi.org/10.21775/cimb.040.081). As the majority of phage clinical application state to actual number of phages administered, in vitro phage concentrations in PFU/ml can be more easily translated in phage applications [22]. The actual phage concentrations are in line with the inundation threshold, which is the minimum phage concentration required to reduce bacterial load [23].

In conclusion, the microbroth growth inhibition assay correlated well with the standard host-range studies in overlay agar method demonstrating 100% sensitivity and specificity in detecting phages with any lytic activity with 6 h of incubation. It is more subjective, provides quantitative data, and is high-throughput, as several phages and bacterial strains can be tested at the same time.