My group has published scientific papers showing that vaccination of guinea pigs with the novel, antibiotic-less version of BCGΔBCG1419, protected against pulmonary and extrapulmonary tuberculosis (TB) better than its parental BCG, produced upon infection with M. tuberculosis H37Rv (https://doi.org/10.1038/s41598-021-91993-8), and that it also improved protection against pulmonary TB produced upon infection with the clinical strain M. tuberculosis M2, while it induced improved T CD8+ memory cell responses (https://doi.org/10.1038/s41598-022-20017-w). It is worth noting that for these studies, we used lab-scale produced vaccines using so called “standard conditions” (e.g., mid-log phase, planktonic cultures in Middlebrook 7H9 OADC Tween 80, or Proskauer-Beck Tween 80, shaken). I would like to point out that most recent preclinical studies of novel TB vaccine candidates employ bacteria harvested at mid-log phase, using planktonic cultures in Middlebrook 7H9 OADC Tween 80.
However, production process followed by manufacturers are referred to occur in cultures as surface pellicles, in Sauton media (or an undisclosed media or even manufacturer-owned and/or protected method), therefore, safety and efficacy upon production in these conditions (Sauton media, surface pellicles) for experimental vaccine candidates remaining largely unknown. This in turn poses a potential translational problem: How different in its safety, immunogenicity and efficacy profiles would a vaccine candidate produced and tested in one growth condition, optimized for dispersing cells (e.g. Middlebrook 7H9 with ADC/OADC plus detergent -such as Tween 80- or any other agent suitable for this purpose) would be from the same vaccine candidate once that is escalated and produced under different conditions (say, for instance, Sauton media with no detergent, as surface pellicles)? Other potential translational problems are: (a) How easy would it be for vaccine manufacturers to transition from surface pellicles to shaken-produced (e.g., bioreactors, fermentors, etc.) of current BCG or any other novel TB vaccine candidate? (b) How willing vaccine manufacturers would be to indeed transition their processes? (c) Would there be enough investment to determine in clinical trials, the safety and immunogenicity of these shaken-produced BCG or other novel vaccine candidates?
I acknowledge that growth of BCG as surface pellicles and bioreactors in Sauton medium resulted in similar protection against M. tuberculosis challenge in mice (https://doi.org/10.1016/S0168-1656(02)00046-9 ), however, these experiments were conducted by using vaccination and challenge both by intravenous routes, which, for the former cannot be employed in humans, and for the latter, it does not resemble the aerosol transmission model of TB.
Growth conditions employed to produce BCG is one of many potential factors contributing to variable efficacy of protection (https://doi.org/10.3390/vaccines10010057), proven to be the case in a murine model of TB (https://doi.org/10.1016/j.vaccine.2011.12.044), and it may also contribute to the observed changes in viability, RNA content and capacity to induce ex vivo immune responses recently reported for BCG-Denmark (DEN), -Japan (JPN), -India (IND), -Bulgaria (BUL) and -USA (https://doi.org/10.1016/j.vaccine.2019.11.060).
In summary, I think there is the need to start addressing the questions posed above as potential translational problems, at the preclinical level to start with. Towards this end, I think the easiest and at hand solution would be that novel TB vaccine candidates based on live, attenuated mycobacteria should shift the way they are produced for evaluation of their safety, immunogenicity and efficacy, from mid-log phase, planktonic cultures in Middlebrook 7H9 OADC Tween 80 to surface pellicles produced in Sauton media.
This of course would pose potential limitations in effectively estimating vaccine dose (one of the main reasons for dispersing BCG and M. tuberculosis cultures used in research) yet as long as this continues to be the main condition employed by vaccine manufacturers, it may be worthwhile to change the way we estimate vaccine dose from viable colony-forming units (CFU) to either protein content, or still maintain assessment of viability but not based on CFU numbers but rather by determining it by flow cytometry, or by assessing bacterial metabolic activity or even rate of transcriptional activity.