Surveillance of Daughter Micronodule Formation Is a Key Factor for Vaccine Evaluation Using Experimental Infection Models of Tuberculosis in Macaques

Tuberculosis (TB) is still a major worldwide health problem and models using non-human primates (NHP) provide the most relevant approach for vaccine testing. In this study, we analysed CT images collected from cynomolgus and rhesus macaques following exposure to ultra-low dose Mycobacterium tuberculosis (Mtb) aerosols, and monitored them for 16 weeks to evaluate the impact of prior intradermal or inhaled BCG vaccination on the progression of lung disease. All lesions found (2553) were classified according to their size and we subclassified small micronodules (<4.4 mm) as ‘isolated’, or as ‘daughter’, when they were in contact with consolidation (described as lesions ≥ 4.5 mm). Our data link the higher capacity to contain Mtb infection in cynomolgus with the reduced incidence of daughter micronodules, thus avoiding the development of consolidated lesions and their consequent enlargement and evolution to cavitation. In the case of rhesus, intradermal vaccination has a higher capacity to reduce the formation of daughter micronodules. This study supports the ‘Bubble Model’ defined with the C3HBe/FeJ mice and proposes a new method to evaluate outcomes in experimental models of TB in NHP based on CT images, which would fit a future machine learning approach to evaluate new vaccines.


Introduction
Understanding the dynamics of tuberculosis (TB) disease is essential for the design of better strategies against it, as TB is a cause of high morbidity and mortality worldwide. Mycobacterium tuberculosis (Mtb) infects about 10 million people every year and a quarter of humanity is estimated to be infected. Moreover, it has been the deadliest single infectious disease for decades, killing nearly 1.5 million people annually [1]. Unfortunately, this mortality has been overtaken by SARS-CoV-2 in 2020, a further consequence of which has been a devastating impact on TB care resulting in an estimated additional half a million TB this model was built on data obtained from the murine model with C3HeB/FeJ mice by Marzo et al. [54]. Supportive data obtained from the NHP model would endorse future confirmation in humans.
Image analysis plays a fundamental role in TB disease assessment. Chest X-ray was classically used for diagnosis and still remains a basic tool due to its widespread availability and sufficient sensitivity. Recently, high-quality diagnostic images such as computed tomography (CT), positron emission tomography-computed tomography (PET-CT) and magnetic resonance imaging (MRI) have emerged. However, to our knowledge, there have been limited opportunities to characterize the development of Mtb-induced disease, particularly the early events after infection. Scanning enables the collection of high-quality data in a non-invasive way, allowing the examination of features of particular interest such as lung consolidations, nodules, micronodules (daughter and isolated), cavitation, tree-in-bud lesions and their distribution [5,11,46,48,49,[54][55][56][57][58][59][60][61][62]. CT scanning allows the acquisition of high-quality images in a few seconds (notably faster than MRI and PET-CT), thus minimizing interference caused by respiratory motion, and being also the most available and economical three-dimensional imaging technology. A score system for the estimation of TB burden disease in the NHP model based on CT images developed by Sharpe et al. [63] has been reported.
The aim of this study was to evaluate the progression of Mtb infection in experimental NHP models through the spatial monitoring of the micronodules. We found that the identification of consolidations associated with daughter micronodules would be a useful tool to evaluate the efficacy of new vaccines.

Animals
Images were obtained from 35 macaques aged 3-4 years; 13 cynomolgus macaques (Macaca fascicularis, of Indonesian genotype) and 22 rhesus macaques (Macaca mulatta, of Indian genotype). The macaques enrolled in the studies were obtained from established closed breeding colonies in the United Kingdom and absence of previous exposure to mycobacterial antigens was confirmed by tuberculin skin test as part of colony management procedures and by screening for IFN-γ ELISPOT (MabTech, Nacka, Sweden) responses to tuberculin-PPD (SSI, Copenhagen, Denmark), and pooled 15-mer peptides of ESAT6 and CFP10 (Peptide Protein Research LTD, Fareham, UK). BCG vaccinations were delivered to sedated animals either as a 100 µL intradermal (ID) injection using Danish strain 1331 (SSI, Copenhagen, Denmark) delivered to the upper left arm, or by exposure to aerosolised BCG Danish strain 1331 created using an Omron MicroAir mesh nebuliser (Omron Healthcare UK Ltd., Milton Keynes, UK) and a modified paediatric anaesthesia mask. The BCG vaccine was prepared for intradermal administration according to manufacturer's instructions for administration to humans. One ml of Sautons diluent was added to a vial of BCG vaccine to give a suspension of BCG at an estimated concentration of between 2 and 8 × 10 6 CFU/mL. The vaccination dose was selected to be equivalent to a standard adult intradermal dose [13,14,28]. The macaques were experimentally infected with Mtb (Erdman strain K 01) ultra-low dose aerosol exposure (presented dose 19-26 CFU) as previously described [14], and following exposure to MTB, macaques were housed in facilities compliant with the Advisory Committee for Dangerous Pathogens Level 3 (ACDP3) regulations. Compatible social groups were housed in accordance with the Home Office (UK) Code of Practice for the Housing and Care of Animals Used in Scientific Procedures (1989), and the National Committee for Refinement, Reduction and Replacement (NC3Rs) Guidelines on Primate Accommodation, Care and Use, August 2006(NC3Rs, 2006. Cages were approximately 2.5 M high by 4 M long by 2 M deep, constructed with high level observation balconies and with a floor of deep litter to allow foraging. Additional environmental enrichment was afforded by the provision of toys, swings, feeding puzzles and DVDs for visual stimulation. In addition to standard old-world primate pellets, diet was further supplemented with a selection of fresh vegetables and fruit. For each procedure, sedation was applied by intramuscular injection with ketamine hydrochloride (10 mg/kg) (Ketaset, Fort Dodge Animal Health Ltd., Southampton, UK). None of the animals had been used previously for experimental procedures and each socially compatible group was randomly assigned to a particular study treatment. The study design and all procedures were approved by the UKHSA Porton Down Animal Welfare and Ethical Review Body, and authorised under an appropriate UK Home Office project license.
Animals were sedated at regular intervals to measure body weight and temperature, red blood cell (RBC) haemoglobin levels and erythrocyte sedimentation rate (ESR) and for collection of blood samples, X-radiographs or CT scans. Animal behaviour was observed throughout the studies for contra-indicators and the time of necropsy, if prior to the end of the planned study period, was determined by experienced primatology staff based on a combination of the following adverse indicators: depression or withdrawn behaviour, abnormal respiration (dyspnoea), loss of 20% of peak post-challenge weight, ESR levels elevated above normal (>20 mm), haemoglobin level below normal limits (<100 g/dL), increased temperature (>41 • C) and abnormal thoracic radiograph.
The studies reported describe individuals with active TB that was either controlled during the study period or progressed to levels that met humane endpoint criteria. TB was not latent in any of the subjects.

Experimental Groups
Once CT images analysis was concluded, we defined five experimental groups among the thirty-five macaques, based on the genotype and their vaccination status: cynomolgus unvaccinated (CUV, n = 10), cynomolgus intradermal-BCG-vaccinated (CIDV, n = 3), rhesus unvaccinated (RUV, n = 11), rhesus aerosolized, or inhaled, BCG-vaccinated (RAEV n = 6) and rhesus intradermal-BCG-vaccinated (RIDV, n = 5). These groups together with CT scan schedule are shown in Figure 1 and Supplementary Figure S1. Macaques received BCG vaccination (green circles) delivered by aerosol or intradermal injection, or were left as unvaccinated controls. All animals received ultra-low dose aerosol challenge with Mtb 21 weeks after. Gray triangles indicate application of 'in vivo' CT scanning. Light grey triangles indicate the time points where not all animals were scanned and the number between brackets shows in how many of them this was effectively performed. In blue, unvaccinated cynomolgus (CUV). In cyan, intradermal-vaccinated cynomolgus (CIDV). In red, unvaccinated rhesus (RUV). In orange, aerosol-vaccinated rhesus (RAEV). In purple, intradermal-vaccinated rhesus (RIDV).

CT Images
We analysed 123 computed tomography (CT) scans obtained from subjects enrolled in different studies [13,14,28] where sequential images were acquired from each macaque on three or four occasions during the post-infection period from the 3rd to the 16th week ( Figure 1 and Supplementary Figure S1). Chest multidetector CT scans were obtained by using a 16 slice Lightspeed CT scanner (General Electric Healthcare, Milwaukee, WI, USA), from sedated subjects while free breathing. Technical parameters applied: tube voltage, 120 kVp; tube current modulation, 100-150 mA. Reconstructions were made using a high spatial frequency algorithm and lung window at a slice thickness of 0.625 mm. These technical parameters are extracted from the original animal experimental studies [13,14,28].

TB Lung Lesions Analysis from CT Images
CT scans were evaluated by a medical consultant radiologist with expertise in respiratory diseases and her supervisor. Both were blinded to vaccination status and clinical data. Data show the consensus between both. Images were then examined in Philips IntelliSpace Portal software (2015 Koninklijke Philips N.V.) by multiplanar reformatting axial, coronal and sagittal, and maximum intensity projection (MIP) to detect the smallest nodules. The Tumor Tracking tool from the mentioned software was used to analyse the volume and main axis of each lesion. We used standard image descriptors, accepted and reproducible concepts from radiologic lexicon detailed on Fleischner Society Glossary as well as in previous studies based on CT images [63].
A. Micronodule is a small solid lung nodule with smooth margins. B. Consolidation (also referred to as consolidated lesion) is a pneumonic patch or necrotizing consolidative process that occupies and even destroys alveoli, and it is described in CT images as a soft tissue lesion within lung parenchyma showing irregular margins. C. Cavitation corresponds to the appearance of gas within lung consolidations. It is described as a central radiolucency image, which matches with gas density, surrounded by soft tissue [8,57]. D. Pleural distance is defined as the distance between the margin of each consolidation to the closest pleura or fissure.

Statistical Assessment
To compare the mean and variance of the different observations in the two groups we used the one-way ANOVA test and Kruskal-Wallis test to determine the existence of significant differences. To compare the observed distribution of data into discrete categories, we built a contingency table and the Fisher's test. Implementation was conducted in MATLAB and Prism. We computed the mean with available data. In the plots, we indicated with a dashed line if there were missing data. Recollected data are available in the Supplementary Material (Supplementary File raw data).

Preliminary Images' Assessment and Anatomy
We were provided with high-quality CT scan images, similar to those acquired for medical diagnosis in humans. A first anatomical examination corroborated that cynomolgus macaques had a smaller lung volume than rhesus macaques (Supplementary Figure S2). Overall, 323 consolidations were identified; 285 were found in rhesus and 38 in cynomolgus, with a median of 14 and 2 lesions per animal, respectively (Table 1). A total of 2228 micronodules were detected, the majority of which (1924) were found in rhesus macaques, while 304 micronodules were detected in cynomolgus ( Table 2). Collected data revealed that lesions globally increase over time. Right lung lobes developed more consolidations in number but left lung lobes presented with a higher proportion occupied by disease. A non-significant trend was seen for the major involvement to occur in the upper lobes, which was higher in the left side (Supplementary Figure S3). Further CT scans' inspection showed the lung lobes to be separated by pleural fissures as well as bronchovascular bundles, thus presenting the same airway anatomy pattern seen in humans. The right lung is the largest and is composed of upper, middle, lower and infracardiac (or accessory or azygous) lobes, and the left lung is composed of upper, middle (or lingula) and lower lobes [64][65][66][67][68][69].
However, in our study, the fissure for the middle left lobe was frequently incomplete or even absent.
We established the following measurements per each lesion: For micronodules, we considered all lesions with a maximum axis ranging from 1 to 4.4 mm (which equates to a maximum volume of 0.044 cm 3 ). Their length mainly falls in the granuloma's size (1-3 mm), but we also included rounded spherical lesions with a maximum axis under 4.5 mm, as they tended to be stable. In contrast, almost all lesions above 4.5 mm increased or showed an instable size. We defined two subgroups: (a) Daughter micronodules, found around consolidations, such as satellites, that are presumably generated from that consolidated lesion [5,49,54]. They are located within an annular zone, where the amplitude of this circular crown corresponds to the consolidation's short axis; they were counted for each CT scan ( Figure 2). (b) Isolated micronodules, the ones independent or not associated with a consolidated lesion ( Figure 3). The size and number of them were determined together within their bronchopulmonary segment.
Consolidations are large enough to be followed up and carefully tracked over consecutive scans, and their axis (mm) and volume (cm 3 ) were measured ( Figure 4). We considered that these kinds of lesions modify their size over time and are equal or bigger than 4.5 mm. disease. A non-significant trend was seen for the major involvement to occur in the upper lobes, which was higher in the left side (Supplementary Figure S3). Further CT scans' inspection showed the lung lobes to be separated by pleural fissures as well as bronchovascular bundles, thus presenting the same airway anatomy pattern seen in humans. The right lung is the largest and is composed of upper, middle, lower and infracardiac (or accessory or azygous) lobes, and the left lung is composed of upper, middle (or lingula) and lower lobes [64][65][66][67][68][69]. However, in our study, the fissure for the middle left lobe was frequently incomplete or even absent.
We established the following measurements per each lesion: For micronodules, we considered all lesions with a maximum axis ranging from 1 to 4.4 mm (which equates to a maximum volume of 0.044 cm 3 ). Their length mainly falls in the granuloma's size (1-3 mm), but we also included rounded spherical lesions with a maximum axis under 4.5 mm, as they tended to be stable. In contrast, almost all lesions above 4.5 mm increased or showed an instable size. We defined two subgroups: (a) Daughter micronodules, found around consolidations, such as satellites, that are presumably generated from that consolidated lesion [5,49,54]. They are located within an annular zone, where the amplitude of this circular crown corresponds to the consolidation's short axis; they were counted for each CT scan ( Figure 2). (b) Isolated micronodules, the ones independent or not associated with a consolidated lesion ( Figure 3). The size and number of them were determined together within their bronchopulmonary segment.
Consolidations are large enough to be followed up and carefully tracked over consecutive scans, and their axis (mm) and volume (cm 3 ) were measured ( Figure 4). We considered that these kinds of lesions modify their size over time and are equal or bigger than 4.5 mm.      Cavitations must have a greater size than the surrounding bronchial calibre, for avoiding misinterpretation with air bronchogram. It was assessed qualitatively, being present (yes) or absent (no).
In pleural distance measurements we also determined the lung volume from each animal, using the COPD tool from the Philips IntelliSpace Portal software, which allowed the evaluation of the main airway and volume from both lungs separately.
We also observed foci of bronchocentric pneumonia, detected as a 'tree-in-bud' pattern in CT images, which is a common CT finding in active lung TB [58][59][60][61]70,71]. However, we did not include them in our analyses due to their lack of homogeneity and instability, which made them difficult to demarcate and to follow-up.
RUV05 presented with miliary tuberculous disease and developed hundreds of lung micronodules distributed diffusely and homogeneously (as well as extrapulmonary miliary disease). Such profuse dissemination is well known and has been described in other studies as 'widespread discrete pulmonary nodules' [63,72]. It is uncommon and denotes extremely uncontrolled infection with hematogenous dissemination. Our study was focused on pulmonary pneumonia and airway dissemination; hence, we removed the data collected from this animal from the statistical evaluation due to the different TB disease. Cavitations must have a greater size than the surrounding bronchial calibre, for avoiding misinterpretation with air bronchogram. It was assessed qualitatively, being present (yes) or absent (no).
In pleural distance measurements we also determined the lung volume from each animal, using the COPD tool from the Philips IntelliSpace Portal software, which allowed the evaluation of the main airway and volume from both lungs separately.
We also observed foci of bronchocentric pneumonia, detected as a 'tree-in-bud' pattern in CT images, which is a common CT finding in active lung TB [58][59][60][61]70,71]. However, we did not include them in our analyses due to their lack of homogeneity and instability, which made them difficult to demarcate and to follow-up.
RUV05 presented with miliary tuberculous disease and developed hundreds of lung micronodules distributed diffusely and homogeneously (as well as extrapulmonary miliary disease). Such profuse dissemination is well known and has been described in other studies as 'widespread discrete pulmonary nodules' [63,72]. It is uncommon and denotes extremely uncontrolled infection with hematogenous dissemination. Our study was focused on pulmonary pneumonia and airway dissemination; hence, we removed the data collected from this animal from the statistical evaluation due to the different TB disease.

Mtb-Induced Disease Is Worse in Rhesus Macaques
In order to exemplify the complexity of the operational procedure, Figure 5 illustrates a particular evolution in one case (RIDV02). Overall development of the natural infection in control groups is summarized in Figure 6. The temporal evolution of the lesions is shown per animal, based on the size and localization per lobe in a 2D representation. Each pair of lungs is reshaped according to each individual volume (adapted to the image obtained from macaque RUV04).
At first sight, this summary highlights the greater severity of the infection in rhesus. In particular, at week 3, cynomolgus developed a low number of lesions, with a small volume, which were usually well controlled and exhibited minimal enlargement with time. Lesions were not identified in one individual (CUV10), although infection was confirmed by other methodologies. Other individuals showed a decline in condition only at the late phase (15th week) of the study, e.g., CUV07, in which an extraordinary growth of consolidations was detected on CT images that was probably due to total main bronchus stenosis (despite it being uncommon in cynomolgus).
In contrast, lesions in rhesus were abundant and larger at week 3, usually exhibiting enlargement and dissemination over time. Poor control leading to extreme disease was evident in RUV05 (which developed profuse miliary dissemination, described above) and also in RUV06 and RUV08; these animals had to be euthanized before finishing the monitoring period due to progressive disease that met humane endpoint criteria. However, there was at least one case in a rhesus where a spontaneous weak regression of lesions was observed (RUV10) ( Figure 6).
Cynomolgus macaques developed significantly fewer lung consolidations and micronodules, resulting in a smaller pulmonary volume occupation than that seen in rhesus macaques, as shown in Table 1 and Figure 7. Cynomolgus also had a significantly lower percentage of daughter micronodules than rhesus (Tables 1 and 2, and Figure 8). The overall ratio of daughter/isolated micronodules for non-vaccinated macaques comprising all time points was 1.32 and 0.34 for rhesus and cynomolgus, respectively. Similarly, the number of consolidated lesions associated with daughter micronodules was also significantly lower in cynomolgus (Table 3). These data reflect a higher containment of lesions' progression in cynomolgus.
The size of consolidated lesions was highly dependent on the number of daughter micronodules. This is reflected in Figure 9 where a clear and significant correlation is noted between the size of the consolidated lesions and the number of daughter micronodules, regardless of the vaccination status. This positive slope is markedly higher in rhesus. In addition, a greater portion of occupied pulmonary volume is correlated to a higher number of isolated micronodules (Supplementary Figure S4), so the higher the infectious involve-Pathogens 2023, 12, 236 9 of 21 ment of lungs, the higher the rate of pulmonary dissemination (local and endobronchial reinfections, which fits with the Dynamic Hypothesis [52]). there was at least one case in a rhesus where a spontaneous weak regression of lesions was observed (RUV10) ( Figure 6). week shows a higher enlargement of the consolidation, in which arose a cavitation (red arrow).
Moreover, some small isolated micronodules (green arrowheads) appeared distributed on both lungs. Notice that nodular images not labelled correspond to pulmonary vessels.  Moreover, some small isolated micronodules (green arrowheads) appeared distributed on both lungs. Notice that nodular images not labelled correspond to pulmonary vessels. cronodules, resulting in a smaller pulmonary volume occupation than that seen in rhesus macaques, as shown in Table 1 and Figure 7. Cynomolgus also had a significantly lower percentage of daughter micronodules than rhesus (Tables 1 and 2, and Figure 8). The overall ratio of daughter/isolated micronodules for non-vaccinated macaques comprising all time points was 1.32 and 0.34 for rhesus and cynomolgus, respectively. Similarly, the number of consolidated lesions associated with daughter micronodules was also significantly lower in cynomolgus (Table 3). These data reflect a higher containment of lesions' progression in cynomolgus.
.      . Figure 9. Correlation between the volume of consolidated lesions and the number of daughter micronodules. Light circles represent a single consolidation. Dark circles represent the mean volume value of all consolidations that present the same number of daughter micronodules. Correlation value between both quantities is written as r. The significance of correlation value is determined from the p-value that is shown under correlation value, according to the X test. In blue, unvaccinated cynomolgus (CUV); in cyan, intradermal-vaccinated cynomolgus (CIDV); in red, unvaccinated rhesus (RUV); in orange, aerosol-vaccinated rhesus (RAEV); in purple, intradermal-vaccinated rhesus (RIDV); in grey, all macaques (all).
We also detected some transitional lesions, which correspond to isolated micronodules that increased to a size of more than 4.4 mm from one CT to the next, thus becoming a consolidation (Supplementary Table 1). This enlargement occurred mainly in the early phase between the third and eighth week post-challenge. Remarkably, rhesus developed directly a large number of consolidations in the early phase after infection, in contrast to cynomolgus; hence, transitional lesions represent a higher impact on cynomolgus. Importantly, there is no significant difference between both macaque genotypes considering the progression of isolated micronodules towards consolidated lesions.

BCG Vaccination in Rhesus Macaques Reduces the Number of Lesions, although Intradermal Vaccination Better Controls Daughter Micronodules
Initial evaluation of the impact of vaccination in macaques suggested that a protective effect appeared by the 3rd week post-challenge (Figures 6, 8 and 10). However, a We also detected some transitional lesions, which correspond to isolated micronodules that increased to a size of more than 4.4 mm from one CT to the next, thus becoming a consolidation (Table 1). This enlargement occurred mainly in the early phase between the third and eighth week post-challenge. Remarkably, rhesus developed directly a large number of consolidations in the early phase after infection, in contrast to cynomolgus; hence, transitional lesions represent a higher impact on cynomolgus. Importantly, there is no significant difference between both macaque genotypes considering the progression of isolated micronodules towards consolidated lesions.

BCG Vaccination in Rhesus Macaques Reduces the Number of Lesions, Although Intradermal Vaccination Better Controls Daughter Micronodules
Initial evaluation of the impact of vaccination in macaques suggested that a protective effect appeared by the 3rd week post-challenge (Figures 6, 8 and 10). However, a deeper analysis revealed that protection measured in terms of infection (i.e., occupied lung fraction at week 3), did not appear to provide a reliable indicator of later outcome ( Supplementary  Figures S5 and S6). When each disease volume fraction was normalized against the maximum infiltrated lung volume per macaque (Figure 11), it revealed a clear impact of the vaccination in cynomolgus macaques early after challenge, although it did not last until the end of the monitoring time.   Lesions (consolidations and micronodules), in black, are represented as spheres of the measured volume and located in the segment where they were identified. (A) In cyan, the reconstructions of lungs from intradermal-vaccinated cynomolgus (CIDV) for each CT scan. (B) In orange, the reconstructions of lungs from aerosol-vaccinated rhesus (RAEV) for each CT scan. In purple, the reconstructions of lungs from intradermal-vaccinated rhesus (RIDV) for each CT scan. Macaque RAEV04 marked with dagger symbol was euthanized before the planned end of study, according to the welfare monitoring plan.  Due to the correlation between the volume of the consolidated lesions and the number of daughter micronodules, we investigated if vaccination had an impact on this fraction of micronodules in relation to the enlargement of consolidated lesions, and this was the case. Looking at Figure 8, vaccination clearly reduced this fraction in all vaccinated macaques, although temporally in rhesus, as it rapidly resumed by week 7 to 8. The reduction in this parameter only prevails in cynomolgus until the 15th week.
Evaluation of the route of vaccination in rhesus revealed an interesting paradox. While aerosol vaccination (RAEV) reduced the total number of micronodules, more than intradermal vaccination (RIDV), a significantly greater fraction of daughter micronodules were present (3.03) ( Table 4). Thus, the protective effect of RAEV relies on the reduction in the total number of micronodules (Tables 2 and 3) rather than the control of subsequent consolidation development. Even though the reduction in total micronodules was weaker in the RIDV, its effect was noted in both types of micronodules, even when this fraction was slightly higher in RIDV (1.51) than non-vaccinated (RUV) (1.32) ( Table 2). At the end, this means that progression towards consolidation at a hypothetical later time point would be faster in the RAEV group, followed by RIDV and RUV.
The effect of vaccination in rhesus was also observed as a decrease in the proportion of consolidated lesions that were associated with daughter micronodules, which was more pronounced and significant after intradermal (95%) than after aerosol vaccination (99%) ( Tables 3 and 5). This effect was even more pronounced in cynomolgus, although in this case there was no variation in the proportion of isolated and daughter micronodules between vaccinated (intradermal) and unvaccinated groups (Tables 2 and 4). This can be explained because, as indicated above, cynomolgus are able to spontaneously stop the progression of the consolidated lesions soon after the infection (i.e., before week 3 after challenge).

Cavitation Is More Probable in Larger Lesions and Contact with Pleura Appear to Impact on TB Progression
Cavitation occurred mostly in rhesus and especially at the late phase, where 34.4% of consolidated lesions developed cavities in non-vaccinated rhesus (RUV), closely followed by 28.4% of the aerosolized (RAEV) and by 24.3% of the ID (RIDV). Analysis of cavitation in rhesus ( Figure 12) revealed a relationship with lesion size (p < 0.0001): the higher the consolidation volume, the higher the rate of cavitation, with cavitation more probable in lesions greater than 2 cm 3 , which represents a rate of 57%. In contrast, almost 80% of lesions from 0 to 2 cm 3 did not cavitate. The anatomic distribution of lesions was not found to influence cavity formation.
We also evaluated the relationship between consolidations and pleura, in order to investigate if the contact with pleura favoured the encapsulation process of the lesions and stopped enlargement. As shown in Supplementary Figure S7, the group of smaller lesions (up to 0.5 cm 3 , which accounts for the higher number of consolidations) were mainly not in contact with the pleura, especially in rhesus (which have bigger lungs), and from this size on, consolidations were mainly in contact with the pleura. The next step was to determine if consolidations changed between two consecutive CT scans, with respect to their size and distance from pleura. Thus, we evaluated the change in the volume together with the relationship with the contact with pleura. We explored the change in the relative volume gradient, explained through a boxplot ( Figure 13). In this regard, differences are significant (p = 0.04 Kruskal-Wallis) in cynomolgus for those consolidations that make contact with the pleura, as they remain stable in volume. Thus, pleural contact is protective in cynomolgus, preventing the progression of consolidations, and this trend is irrespective of their vaccination status. In contrast, in rhesus, consolidations that enlarge and make contact with pleura do not experience a volume control because the increase in the size is too fast and large to benefit from the encapsulation process activated by the pleura. consolidated lesions developed cavities in non-vaccinated rhesus (RUV), closely followed by 28.4% of the aerosolized (RAEV) and by 24.3% of the ID (RIDV). Analysis of cavitation in rhesus ( Figure 12) revealed a relationship with lesion size (p < 0.0001): the higher the consolidation volume, the higher the rate of cavitation, with cavitation more probable in lesions greater than 2 cm 3 , which represents a rate of 57%. In contrast, almost 80% of lesions from 0 to 2 cm 3 did not cavitate. The anatomic distribution of lesions was not found to influence cavity formation.
. We also evaluated the relationship between consolidations and pleura, in order to investigate if the contact with pleura favoured the encapsulation process of the lesions and stopped enlargement. As shown in Supplementary Figure 7, the group of smaller lesions (up to 0.5 cm 3 , which accounts for the higher number of consolidations) were mainly not in contact with the pleura, especially in rhesus (which have bigger lungs), and from this size on, consolidations were mainly in contact with the pleura. The next step was to determine if consolidations changed between two consecutive CT scans, with respect to their size and distance from pleura. Thus, we evaluated the change in the volume together with the relationship with the contact with pleura. We explored the change in the relative volume gradient, explained through a boxplot ( Figure  13). In this regard, differences are significant (p = 0.04 Kruskal-Wallis) in cynomolgus for those consolidations that make contact with the pleura, as they remain stable in volume. Thus, pleural contact is protective in cynomolgus, preventing the progression of consolidations, and this trend is irrespective of their vaccination status. In contrast, in rhesus, consolidations that enlarge and make contact with pleura do not experience a volume control because the increase in the size is too fast and large to benefit from the encapsulation process activated by the pleura. .

Discussion
TB remains a major killer of humankind, and the lack of a surrogate of protection, means there is a constant need for the field to find experimental models able to elucidate new biomarkers and correlates to assist the development of new more effective vaccines. Nowadays, the macaque is considered the most clinically relevant experimental infection model of TB given its physiological closeness with humans and the similarity in disease manifestations. However, the macaque does lack the interlobular septa, present in human lungs, which have been demonstrated to provide a protective mechanism in the minipig

Discussion
TB remains a major killer of humankind, and the lack of a surrogate of protection, means there is a constant need for the field to find experimental models able to elucidate new biomarkers and correlates to assist the development of new more effective vaccines. Nowadays, the macaque is considered the most clinically relevant experimental infection model of TB given its physiological closeness with humans and the similarity in disease manifestations. However, the macaque does lack the interlobular septa, present in human lungs, which have been demonstrated to provide a protective mechanism in the minipig model [9]. The outcome of experimental infection with Mtb has been widely studied in macaques, and models established in rhesus and cynomolgus have been used to assess the efficacy of new drugs and vaccines. Mtb infection in cynomolgus resembles what occurs in humans, with the display of the whole spectrum of human TB, mainly controlling the infection and becoming latently infected. In contrast, rhesus experience a rapid dissemination of the infection in the lung, with the development of active TB in almost all the infected animals, a characteristic that enables the capacity of vaccines to protect against this progression to be assessed using manageable group sizes [73].
There are little data on vaccines showing significant protection against Mtb challenge in the macaque model with reports limited to BCG delivered intravenously (IV), for which the clinical deplorability is debatable, or a CMV-vectored vaccine [74] that is at the preclinical stage of development, and MTBVAC [28]. While there are very little data on immune responses induced in NHP and humans after exactly the same immunizations, the immune profiles induced following immunization with MTBVAC were shown to reflect those identified in human clinical trials [75][76][77]. This concordance between immune profiles measured in clinical trials and a macaque preclinical assessment of a novel TB vaccine candidate, demonstrating significantly improved outcomes after Mtb challenge, is a promising indication that the protection provided by MTBVAC vaccination to macaques will translate to the human population.
The ability to demonstrate vaccine efficacy is critical, and the role of the macaque models in vaccine development is underpinned by the ability to measure vaccine-associated impacts on the disease that develops following infectious challenge.
Our approach has been to test a new way to interpret data derived from these models based on the 'Bubble Model', that defines the progression from latent to active TB supported by the interpretation of disease progression defined in the studies on the C3HeB/FeJ mice [49,54,78,79]. As with humans, the C3HeB/FeJ mouse model establishes that the development of a large lesion after the Mtb infection requires a strong inflammatory response in the initial lesions, the induction of daughter micronodules around them and the coalescence of all lesions. To evaluate this concept in macaques, we used CT images obtained previously from studies using an experimental model of low-dose aerosol exposure, thus closely resembling what is thought to happen in the natural infectious process, where CT scans were collected at several time points to monitor the evolution of the infection in each animal [13,14,28]. In line with previous reports, three weeks after infection, more large lesions were identified in rhesus compared to cynomolgus, which has been related to the development of a stronger inflammatory response in the former [14,40]. These lead to an exudative lesion, a hallmark of progression towards active TB [51] and a key issue for its development, according to the bubble model.
Thus far, the assessment of new vaccines has been mainly based on the induction of an immune response and a reduction in disease burden measured using pathology scoring, which included a semiquantitative approach of the involvement of lungs and other organs, as well as imaging-based measures through PET-CT [11,45], CT evaluation [28,80], histology-based measures (e.g., granuloma staging and prevalence), together with bacterial burden in organs and changes in clinical parameters. There is quite a lot of experience in using imaging as a tool to evaluate BCG efficacy in animal models, with special interest in NPH. Some studies appeared during the last few years that mainly focused on BCG alternative routes. White et al. [28] analysed CT images in rhesus and used a quantitative score system for the estimation of TB-induced disease burden [63]; Sharpe et al. [81] and recently Sibley et al. [29] used MRI images in ex vivo lungs from macaques to determine pulmonary disease burden applying stereology [46], as well as counting discrete and coalescent lesions; Darrah et al. [72] and DiFazio et al. [82] performed PET-CT images in macaques and mainly evaluated the grade of tuberculous activity by measuring global lung parenchyma FDG avidity, as well as counting lesions (which were called granulomas). Moreover, there are some experiences including other animal models, as Kraft et al. [83] analysed ex vivo lungs from guinea pigs using MRI, and described their imaging appearance, their anatomical distribution and nodule counting.
Previous analysis of these images highlighted the early reduction in micronodules after BGC vaccination [28]. Our work was able to advance deeper in the exploitation of these data by differentiating the nodules related with the consolidated lesions, daughter micronodules, from the isolated ones. According to the 'Bubble model', daughter micronodules emerge from an initial isolated micronodule, increase in size, consolidate creating larger nodules, which in turn develop more daughter micronodules, a process that enlarges the lesion to the point that it becomes visible by chest X-ray, the hallmark for classifying an Mtb infection as active TB [51]. Data from the C3HeB/FeJ model suggests this progression mainly develops at the beginning of the infection when there is still a lower impact of the adaptive immune response. This is characterized by the infiltration of neutrophiles, which allows the extracellular growth of Mtb, fuelling enlargement and allowing the spread of close new infectious foci due to the drainage of infected foamy macrophages, which in turn attracts new neutrophils, causing the development of daughter lesions [54]. This led us to draw a parallel with the NHP model. On the one hand, rhesus develop highly inflammatory lesions with a high neutrophilic infiltration, a process that might be related with the progression associated to the development of daughter micronodules. Conversely, cynomolgus develop less inflammatory and more contained lesions [40], which would link to a reduced capacity to generate daughter micronodules. The lower inflammatory response, and thus the lower capacity to generate daughter cells, allows the fibroblasts to encapsulate the lesion, stopping its progression [9], and explains why the vicinity of the pleura, and its encapsulation capacity, has only a protective effect in cynomolgus.
Our data provide two new measures of pulmonary disease burden induced following infection with Mtb, namely, the ratio daughter/isolated micronodules and the percentage of consolidated lesions linked to daughter micronodules identified through CT in macaques. These measures provided a new tool to evaluate the protective effect of the licenced vaccine (BCG) and revealed differences in disease progression following either intradermal or mucosal (inhaled) administration, even when both routes were able to reduce the number of consolidated lesions and generated a similar reduction in the number of micronodules. While this mucosal aerosol vaccination was better able to reduce the number of isolated micronodules, potentially because of a better capacity to reach and stimulate the immune response in the whole lung, intradermal administration should have an improved ability to reduce the emergence of daughter micronodules, which are the ones responsible for the progression towards active TB.
In summary, we defined a new tool for the evaluation of pulmonary disease induced by Mtb infection, through the description of 'daughter micronodules', which provides a better understanding of the progression of TB and which can be applied for testing new vaccines. This is also amenable for development as a machine learning tool, as nodules are easily identifiable in CT images and can be quantified. Enhanced measures of TB disease burden and progression, such as those reported here, will refine the models used in the development of new vaccines and reduce the number of animals required, consequently providing important scientific and welfare benefits. Equally, the use of data across studies increased the power of the analysis applied and obtained more information without requiring the use of further macaques in line with the principles of the 3Rs.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/pathogens12020236/s1. Figure S1: Experimental groups and CT scan acquisition schedule; Figure S2: Lung size in macaques; Figure S3: Occupied volume fraction per lung lobe; Figure S4: Correlation between the occupied lung fraction and the number of isolated micronodules; Table S1: Summary lesions found; Figure S5: Cynomolgus vaccine effect analysis; Data Availability Statement: Raw data will be available as supplementary data.